{ "complete": true, "report": { "bucket": "vn_S_CART_reg", "label": "Strong Cartoonish Style", "definition": { "id": "vn_S_CART_reg", "category": "talking_style", "head": "vn_S_CART_reg", "pole": "high", "label": "Strong Cartoonish Style", "floor": 4.0, "canonical_dimension_name": "Cartoonish Style" }, "recipe_status": "direct_manual_recipe", "manual_provenance": [ { "id": "rank4/style_s_cart/full-v2-sealed/sealed-selected", "mission": "style_s_cart", "format_status": "CANONICAL", "heldout": true, "qualified": false, "source_cfg": 2.5, "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": 6.253745379447937, "genuineness": 4.1306591844558715, "wer": 0.010914380105912364, "identity": 0.5346257735043765, "voiceclap": 0.04435582051053643, "fitness": 0.5506578620735062, "steering_v2": 0.45703136338821837, "raw_target": 1.549201791882515, "target_cdf": 0.45703136338821837, "family_hit_all": null, "nominal_hit_all": null, "n": 100, "valid_fraction": 1, "by_arm": { "reference": { "n": 50, "ce": 6.256604738235474, "genuineness": 4.126391491889954, "wer": 0.010274778811069134, "identity": 0.5346257735043765, "fitness": 0.550868942816157, "steering_v2": 0.45661402638853377 }, "none": { "n": 50, "ce": 6.2508860206604, "genuineness": 4.1349268770217895, "wer": 0.011553981400755593, "identity": null, "fitness": 0.5504467813308556, "steering_v2": 0.457448700387903 } } } } ], "selected_wording": { "id": "manual-cff870ad7de8", "general": "Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.", "inline": [ "Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.", "Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.", "End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly." ], "cfg": 2.5, "recipe": "P06", "strategy": "local-Gemma-mutation", "timing_syntax": "canonical", "canonical_labels": true, "neutral_event": "keep", "parent": "cfg-2.5", "origin": "exact_frozen_manual_recipe" }, "selection": { "manual_no_adapter": { "config": { "scale": 0.0, "cfg": 1.25 }, "n": 20, "target_z": -0.27470705553307334, "target_raw": 1.324728971719742, "ce": 6.1782899618148805, "genuineness": 4.3137312769889835, "identity": 0.5333853902295231, "wer": 0.008823529411764706, "valid_fraction": 0.95, "identity_delta": -0.03484898488968611, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, "manual_adapter": { "config": { "scale": 0.75, "cfg": 1.25 }, "n": 20, "target_z": -0.13344726416257402, "target_raw": 1.5181409001350403, "ce": 6.163596034049988, "genuineness": 4.255991697311401, "identity": 0.5440780110657215, "wer": 0.002631578947368421, "valid_fraction": 1, "identity_delta": -0.024156364053487776, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, "grid": [ { "config": { "scale": 0.0, "cfg": 1.25 }, "n": 20, "target_z": -0.27470705553307334, "target_raw": 1.324728971719742, "ce": 6.1782899618148805, "genuineness": 4.3137312769889835, "identity": 0.5333853902295231, "wer": 0.008823529411764706, "valid_fraction": 0.95, "identity_delta": -0.03484898488968611, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.0 }, "n": 20, "target_z": -0.3260742276605956, "target_raw": 1.2543973952531815, "ce": 6.189806842803955, "genuineness": 4.275133621692658, "identity": 0.5304808661341667, "wer": 0.007894736842105263, "valid_fraction": 0.95, "identity_delta": -0.03775350898504257, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 5.0 }, "n": 20, "target_z": 0.6839843703746527, "target_raw": 2.6373626351356507, "ce": 5.686638593673706, "genuineness": 3.988632416725159, "identity": 0.30988077046349644, "wer": 0.055744539043687646, "valid_fraction": 0.85, "identity_delta": -0.25835360465571283, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 4.0 }, "n": 20, "target_z": 0.6483398376962366, "target_raw": 2.588558387756348, "ce": 5.782943558692932, "genuineness": 4.00589154958725, "identity": 0.3509417396038771, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": -0.21729263551533223, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 3.0 }, "n": 20, "target_z": 0.38059082815178347, "target_raw": 2.2219582974910734, "ce": 5.9249187231063845, "genuineness": 4.083400630950928, "identity": 0.4229204636998475, "wer": 0.013759889920880633, "valid_fraction": 1, "identity_delta": -0.14531391141936184, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 2.5 }, "n": 20, "target_z": 0.2893554803813881, "target_raw": 2.0970394909381866, "ce": 5.914878726005554, "genuineness": 4.084028208255768, "identity": 0.43171147629618645, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.13652289882302285, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 2.0 }, "n": 20, "target_z": 0.13559571513049704, "target_raw": 1.8865126848220826, "ce": 6.06212842464447, "genuineness": 4.164675307273865, "identity": 0.4350821852684021, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.13315218985080718, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.5 }, "n": 20, "target_z": -0.15615234966772376, "target_raw": 1.4870532542467116, "ce": 6.174715757369995, "genuineness": 4.32502818107605, "identity": 0.5045978464186192, "wer": 0.013157894736842105, "valid_fraction": 1, "identity_delta": -0.06363652870059014, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.25 }, "n": 20, "target_z": -0.13344726416257402, "target_raw": 1.5181409001350403, "ce": 6.163596034049988, "genuineness": 4.255991697311401, "identity": 0.5440780110657215, "wer": 0.002631578947368421, "valid_fraction": 1, "identity_delta": -0.024156364053487776, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.5 }, "n": 20, "target_z": -0.13994139761148372, "target_raw": 1.5092491775751113, "ce": 6.149894952774048, "genuineness": 4.269799017906189, "identity": 0.5275866031646729, "wer": 0.0058823529411764705, "valid_fraction": 1, "identity_delta": -0.040647771954536435, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.0 }, "n": 20, "target_z": -0.3058593769045548, "target_raw": 1.2820754289627074, "ce": 6.209266400337219, "genuineness": 4.269799065589905, "identity": 0.5249286726117134, "wer": 0.013888888888888888, "valid_fraction": 0.95, "identity_delta": -0.04330570250749588, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.0 }, "n": 20, "target_z": -0.3441894515265344, "target_raw": 1.2295941561460495, "ce": 6.15406482219696, "genuineness": 4.2992963790893555, "identity": 0.534526239335537, "wer": 0.005409356725146199, "valid_fraction": 1, "identity_delta": -0.033708135783672336, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 5.0 }, "n": 20, "target_z": 0.7878906274487133, "target_raw": 2.7796303629875183, "ce": 5.63084135055542, "genuineness": 3.9341877460479737, "identity": 0.3238842859864235, "wer": 0.01964546783625731, "valid_fraction": 1, "identity_delta": -0.2443500891327858, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 5.0 }, "n": 20, "target_z": 0.7490234400847529, "target_raw": 2.726413679122925, "ce": 5.64116678237915, "genuineness": 3.969804275035858, "identity": 0.3499856971204281, "wer": 0.063625085999312, "valid_fraction": 0.85, "identity_delta": -0.2182486779987812, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 5.0 }, "n": 20, "target_z": 0.7370117374554487, "target_raw": 2.709967339038849, "ce": 5.601106023788452, "genuineness": 3.9663524985313416, "identity": 0.289265295304358, "wer": 0.06745248538011696, "valid_fraction": 0.85, "identity_delta": -0.2789690798148513, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 4.0 }, "n": 20, "target_z": 0.6258788986998007, "target_raw": 2.557805025577545, "ce": 5.755044889450073, "genuineness": 3.9789044737815855, "identity": 0.3840052606537938, "wer": 0.020034399724802202, "valid_fraction": 1, "identity_delta": -0.18422911446541548, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 4.0 }, "n": 20, "target_z": 0.6014160179716835, "target_raw": 2.5243106186389923, "ce": 5.79386465549469, "genuineness": 4.010912346839905, "identity": 0.367162049934268, "wer": 0.039069487444100444, "valid_fraction": 0.9, "identity_delta": -0.20107232518494128, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 4.0 }, "n": 20, "target_z": 0.5952148268130665, "target_raw": 2.515819990634918, "ce": 5.791481852531433, "genuineness": 4.010912299156189, "identity": 0.39139867667108774, "wer": 0.034865196078431374, "valid_fraction": 0.85, "identity_delta": -0.17683569844812155, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 3.0 }, "n": 20, "target_z": 0.4146972423962338, "target_raw": 2.2686565637588503, "ce": 5.986077308654785, "genuineness": 4.121684443950653, "identity": 0.43860476575791835, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": -0.12962960936129092, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 3.0 }, "n": 20, "target_z": 0.36752930010594115, "target_raw": 2.2040745437145235, "ce": 5.902977061271668, "genuineness": 4.071476173400879, "identity": 0.44105232916772363, "wer": 0.003125, "valid_fraction": 1, "identity_delta": -0.12718204595148563, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 3.0 }, "n": 20, "target_z": 0.36431883457093844, "target_raw": 2.1996787965297697, "ce": 5.990644335746765, "genuineness": 4.132039856910706, "identity": 0.4414555490016937, "wer": 0.002631578947368421, "valid_fraction": 1, "identity_delta": -0.12677882611751556, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 2.5 }, "n": 20, "target_z": 0.3100830002793057, "target_raw": 2.125419467687607, "ce": 5.93971197605133, "genuineness": 4.13611934185028, "identity": 0.4201259730383754, "wer": 0.014005847953216374, "valid_fraction": 0.95, "identity_delta": -0.14810840208083392, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 2.5 }, "n": 20, "target_z": 0.25522459641722556, "target_raw": 2.050307720899582, "ce": 5.995012807846069, "genuineness": 4.109759974479675, "identity": 0.4165861904621124, "wer": 0.0061111111111111106, "valid_fraction": 1, "identity_delta": -0.15164818465709687, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 2.5 }, "n": 20, "target_z": 0.24626466528940874, "target_raw": 2.0380398452281954, "ce": 5.983098816871643, "genuineness": 4.126705312728882, "identity": 0.44296686090528964, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.12526751421391963, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 2.0 }, "n": 20, "target_z": 0.1205078104594056, "target_raw": 1.8658544301986695, "ce": 6.057561373710632, "genuineness": 4.184131050109864, "identity": 0.44695580303668975, "wer": 0.008823529411764706, "valid_fraction": 0.95, "identity_delta": -0.12127857208251953, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 2.0 }, "n": 20, "target_z": 0.1009765773283193, "target_raw": 1.839112401008606, "ce": 6.130236792564392, "genuineness": 4.182875847816467, "identity": 0.4512620210647583, "wer": 0.014562908496732026, "valid_fraction": 1, "identity_delta": -0.11697235405445099, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 2.0 }, "n": 20, "target_z": 0.0513671808340583, "target_raw": 1.7711875587701797, "ce": 6.096083354949951, "genuineness": 4.210804212093353, "identity": 0.4899519011378288, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.07828247398138047, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.25 }, "n": 20, "target_z": -0.1159179662332669, "target_raw": 1.54214189350605, "ce": 6.1238826513290405, "genuineness": 4.257560741901398, "identity": 0.4966232255101204, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.0716111496090889, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.5 }, "n": 20, "target_z": -0.15478516665703143, "target_raw": 1.4889251917600632, "ce": 6.186629772186279, "genuineness": 4.315614056587219, "identity": 0.4699583828449249, "wer": 0.008513931888544891, "valid_fraction": 1, "identity_delta": -0.09827599227428437, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.25 }, "n": 20, "target_z": -0.18325194608494932, "target_raw": 1.449948674440384, "ce": 6.152476286888122, "genuineness": 4.251912307739258, "identity": 0.4756589286029339, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.0925754465162754, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.5 }, "n": 20, "target_z": -0.1961425757102966, "target_raw": 1.4322989135980606, "ce": 6.143342232704162, "genuineness": 4.268543803691864, "identity": 0.4965652599930763, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": -0.07166911512613297, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.0 }, "n": 20, "target_z": -0.3516601535972273, "target_raw": 1.2193653225898742, "ce": 6.20092658996582, "genuineness": 4.321262526512146, "identity": 0.4952498197555542, "wer": 0.008596491228070175, "valid_fraction": 1, "identity_delta": -0.07298455536365508, "identity_pairs": 20, "gates_pass": false, "gate_scope": "development_provisional_20_pairs" } ], "development_standard": [ { "scale": 0.0, "n": 20, "target_z": -0.663134771780793, "target_raw": 0.792896431684494, "ce": 6.209663534164429, "genuineness": 4.3617429494857785, "identity": 0.5682343751192093, "wer": 0.002631578947368421, "valid_fraction": 1, "identity_delta": 0.0, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 0.5, "n": 20, "target_z": -0.6585937538091096, "target_raw": 0.7991139620542527, "ce": 6.181864213943482, "genuineness": 4.247205269336701, "identity": 0.5896647408604622, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.0214303657412529, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 0.75, "n": 20, "target_z": -0.6025390544736621, "target_raw": 0.8758636683225631, "ce": 6.230711603164673, "genuineness": 4.279526841640473, "identity": 0.5522906735539437, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.015943701565265655, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 1.0, "n": 20, "target_z": -0.6555664150245559, "target_raw": 0.8032589733600617, "ce": 6.175311493873596, "genuineness": 4.342914938926697, "identity": 0.5740254119038581, "wer": 0.0059649122807017545, "valid_fraction": 1, "identity_delta": 0.005791036784648896, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" } ] }, "heldout": { "standard": { "n": 50, "target_z": -0.8384375023942975, "target_raw": 0.552873146533966, "ce": 6.266294784545899, "genuineness": 4.309651837348938, "identity": 0.5579182085394859, "wer": 0.01030679465462074, "valid_fraction": 1, "identity_delta": 0.0, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" }, "manual": { "n": 50, "target_z": -0.4728515638331883, "target_raw": 1.053430882692337, "ce": 6.206407051086426, "genuineness": 4.311409173011779, "identity": 0.5486446915566922, "wer": 0.008484848484848484, "valid_fraction": 1, "identity_delta": -0.009273516982793807, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" }, "adapter": { "n": 50, "target_z": -0.4494921895405944, "target_raw": 1.0854143762588502, "ce": 6.218956470489502, "genuineness": 4.306388335227966, "identity": 0.5586121982336044, "wer": 0.011129305477131564, "valid_fraction": 1, "identity_delta": 0.0006939896941184998, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" }, "matched_no_adapter": { "n": 50, "target_z": -0.4728515638331883, "target_raw": 1.053430882692337, "ce": 6.206407051086426, "genuineness": 4.311409173011779, "identity": 0.5486446915566922, "wer": 0.008484848484848484, "valid_fraction": 1, "identity_delta": -0.009273516982793807, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" } }, "adapter_vs_matched_cfg": { "n": 50, "delta_z": 0.023359374292593944, "ci95_z": [ -0.023847654250217492, 0.07382812287778183 ], "language_deltas": { "en": -0.019140627448713286, "de": 0.10703124619089047, "fr": 0.06718750326495104, "nl": -0.04267576580950238, "es": 0.004394515265343871 }, "languages_improved": 3, "clear_gain": false, "delta_ce": 0.012549419403076172, "delta_genuineness": -0.005020837783813476, "delta_wer": 0.002644456992283079, "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": 0.023359374292593944, "ci95_z": [ -0.023847654250217492, 0.07382812287778183 ], "language_deltas": { "en": -0.019140627448713286, "de": 0.10703124619089047, "fr": 0.06718750326495104, "nl": -0.04267576580950238, "es": 0.004394515265343871 }, "languages_improved": 3, "clear_gain": false, "delta_ce": 0.012549419403076172, "delta_genuineness": -0.005020837783813476, "delta_wer": 0.002644456992283079, "note": "Exploratory language-cluster bootstrap; five voices, one fixed reference voice per language. No familywise claim across72targets." }, "manual_vs_standard": { "n": 50, "delta_z": 0.3655859385611091, "ci95_z": [ 0.27906249847635617, 0.46126953833766454 ], "language_deltas": { "en": 0.39316406318019814, "de": 0.22060545861504782, "fr": 0.33193358932871214, "nl": 0.5418945283251481, "es": 0.3403320533564393 }, "languages_improved": 5, "clear_gain": true, "delta_ce": -0.05988773345947265, "delta_genuineness": 0.001757335662841797, "delta_wer": -0.0018219461697722567, "note": "Exploratory language-cluster bootstrap; five voices, one fixed reference voice per language. No familywise claim across72targets." }, "route": { "adapter": false, "scale": 0.0, "cfg": 1.25, "reason": "Use manual-only configuration; no clear qualifying incremental adapter gain" }, "reference_only": true, "updated": 1793043253.9455214, "example": { "text": "Quelqu’un a déplacé la petite chaise bleue. Elle était ici... près de la grande bibliothèque. Nous la remettrons après le déjeuner.", "sentences": [ "Quelqu’un a déplacé la petite chaise bleue.", "Elle était ici... près de la grande bibliothèque.", "Nous la remettrons après le déjeuner." ], "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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.", "language": "fr", "seed": 1702, "scale": 0.0, "cfg": 1.25, "frames": 128, "reference": true } }, "rows": [ { "id": "vn_S_CART_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_CART_reg-es-0", "key": "0075ba2cb912977f44f44a1ef24e0b7b685490e915211b722eb8a8c6d294321b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8f3e534f09b98a13a236c745eb1c2794ee157c86a74ce8ad207ef7a3eb631f38", "generation_seconds_batch": 21.126510394969955, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0003669261932373, "target_z": 0.21874997823365974, "ce": 5.202532768249512, "genuineness": 3.515732765197754, "identity": 0.4344792366027832, "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_CART_reg/0075ba2cb912977f44f44a1ef24e0b7b685490e915211b722eb8a8c6d294321b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "010cd3cd806103d8becf3dd9f5d9023a64866a289621829991e1ed8989b8fad1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5eb8a78659953be3714aa4fcc21acaa1018d5194718a97e3dae18ab23e7f5a8b", "generation_seconds_batch": 21.985435421927832, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9188036918640137, "target_z": 0.15917968409900934, "ce": 6.165581703186035, "genuineness": 4.400654315948486, "identity": 0.24950742721557617, "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_CART_reg/010cd3cd806103d8becf3dd9f5d9023a64866a289621829991e1ed8989b8fad1.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": "012a522a9dfb7c26263d390c70d209447faeb2ed4e57cbe9f5ee540b7dcd1ddf", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "33062a5aa9168f4dd99202126494c2ca6aa8d61d5b654717ec0b3fe28e381f42", "generation_seconds_batch": 14.846530252019875, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7274452447891235, "target_z": -0.7109375163247552, "ce": 6.268836498260498, "genuineness": 4.432034492492676, "identity": 0.6154295206069946, "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_CART_reg/012a522a9dfb7c26263d390c70d209447faeb2ed4e57cbe9f5ee540b7dcd1ddf.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "0152985bcc1d3da701885eca4da604c1bedeb62f4c856823280f68f592827b43", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0adbd1c4cae8c49bc3a7ea13829163694d10e54d81e6bc67b87abd22e417856c", "generation_seconds_batch": 18.84776872396469, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.576505422592163, "target_z": -0.09082031590099067, "ce": 5.824047088623047, "genuineness": 4.432034492492676, "identity": 0.7178854942321777, "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_CART_reg/0152985bcc1d3da701885eca4da604c1bedeb62f4c856823280f68f592827b43.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "01edfe2e861df085f83c0a69ece02bd8c3c5257c398b870e86030403d950ba8f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4c8a1e9810f63bc554fbfa62ef46c5ee3d342972aa7edb600c229a659baab0d9", "generation_seconds_batch": 19.13083522208035, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2250001430511475, "target_z": 0.38281246190890456, "ce": 5.8875885009765625, "genuineness": 4.080576419830322, "identity": 0.19698119163513184, "hypothesis": "Ik heb het kleine doodtje 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_CART_reg/01edfe2e861df085f83c0a69ece02bd8c3c5257c398b870e86030403d950ba8f.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "7f32bf8c2c4ad75abba4f77bb8fd4d60fb5fd64cfba600e735cb9463025b5d79", "generation_seconds_batch": 14.767348385998048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9734721183776855, "target_z": -0.5312500217663403, "ce": 6.149696350097656, "genuineness": 4.231201171875, "identity": 0.7612375617027283, "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_CART_reg/01f5d5f22d9b51f4e674eec92be80d0710e62f5704093e3409bc811236eec3e5.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "020fecb51e966c03144019c4cc3f553dd72e92b8f4830dc446dc32718021e365", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4accb7e8a7c1135bb3fe1f853e4c31423d4a93417def4331e94b3b01add21b2b", "generation_seconds_batch": 17.266092004952952, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.438936710357666, "target_z": 0.5390625707406058, "ce": 6.117925643920898, "genuineness": 4.18099308013916, "identity": 0.3770729899406433, "hypothesis": "The door is still open. I put your goat 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_CART_reg/020fecb51e966c03144019c4cc3f553dd72e92b8f4830dc446dc32718021e365.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "0255fe545b05cddd02df759e20ca2e66b762d431a5993deef1a1f12c9652787a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1133b04787c9be4b5f0127ca0584c01f53c6f7e198b2047025b8ce9080734e44", "generation_seconds_batch": 16.040086375083774, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.970950722694397, "target_z": 0.19726564268515145, "ce": 5.919358730316162, "genuineness": 4.2374773025512695, "identity": 0.17481262981891632, "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_CART_reg/0255fe545b05cddd02df759e20ca2e66b762d431a5993deef1a1f12c9652787a.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "0273e63cac088b837051e1f03a40adb6b77fcd5ca980087c84bd1673e0bab767", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "83551d6070b5c3745cb16d3d5e7d4447fa385040aa31946d8e7bfa14af515e4d", "generation_seconds_batch": 20.43732152704615, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1098566055297852, "target_z": -0.4316406318019813, "ce": 6.24500846862793, "genuineness": 4.494794845581055, "identity": 0.22856250405311584, "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_CART_reg/0273e63cac088b837051e1f03a40adb6b77fcd5ca980087c84bd1673e0bab767.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "02c705834fa9150036f0bf2d30b9302ac4caaa3812eee1584678972fefcfc239", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "10de3ba529e02e7acd0f8fa55997b16423893ccd164aaf4b53d97b8b4a9feafd", "generation_seconds_batch": 21.010634328005835, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7810821533203125, "target_z": 0.05859376904554772, "ce": 6.022613525390625, "genuineness": 4.18726921081543, "identity": 0.6944593191146851, "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_CART_reg/02c705834fa9150036f0bf2d30b9302ac4caaa3812eee1584678972fefcfc239.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "02db664d45f614cdbe14ba3a154659c999b7e56a46066e47ffd36392db5ebc33", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "981abcd4c5de63e90cf67e799498fed02114bec9eb8ee042022455b4a4461723", "generation_seconds_batch": 22.1942776349606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8558071255683899, "target_z": -0.6171874945584149, "ce": 6.308549880981445, "genuineness": 4.13078498840332, "identity": 0.43293339014053345, "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_CART_reg/02db664d45f614cdbe14ba3a154659c999b7e56a46066e47ffd36392db5ebc33.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": "035e7ea9585db4542b876970aaab4de0d4153120ce57b0c0437a6fd24395cac8", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2ebaba6c977460ee399bee0c84ba7453af9fac9bedcf3be3437473b84d65ce36", "generation_seconds_batch": 14.993886140990071, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1178791522979736, "target_z": -0.42578130713664314, "ce": 6.165581703186035, "genuineness": 4.6203155517578125, "identity": 0.6320537328720093, "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_CART_reg/035e7ea9585db4542b876970aaab4de0d4153120ce57b0c0437a6fd24395cac8.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "0363059421b8df56f99f18de1372c0b72123fccb30fd00155b622077a4f37e06", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2996840f9bc930faac4bcdcf630f2f5d158f088bcdc68fd73359f4511d18cc06", "generation_seconds_batch": 16.120895523927175, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.64752459526062, "target_z": 0.6914062309544523, "ce": 5.6532793045043945, "genuineness": 4.118232727050781, "identity": 0.15033385157585144, "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_CART_reg/0363059421b8df56f99f18de1372c0b72123fccb30fd00155b622077a4f37e06.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": "03be5b5bfaf8a7e0482453b3b4e81b96b09556de23ba47e60873afab36f78ce7", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "15cba0508a025000f03b60400dc28042e61d1eaf555a840cfb7ababf524b5692", "generation_seconds_batch": 19.73768907203339, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6953548192977905, "target_z": -0.7343749891168299, "ce": 6.141753673553467, "genuineness": 4.256305694580078, "identity": 0.5344064235687256, "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_CART_reg/03be5b5bfaf8a7e0482453b3b4e81b96b09556de23ba47e60873afab36f78ce7.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "043e857a93f570c59d8e84a26bdb6ec31656e98841efc73e178b16fda6654dd4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3f0dc1893c5ba2b3fea5c714438d05886d8690f1fb32ba1a2cfe572993765064", "generation_seconds_batch": 19.48264504992403, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1929097175598145, "target_z": 0.35937498911682986, "ce": 5.80021858215332, "genuineness": 3.986435890197754, "identity": 0.552094578742981, "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_CART_reg/043e857a93f570c59d8e84a26bdb6ec31656e98841efc73e178b16fda6654dd4.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "046a90d0259f0270802273d9e45f1166ae6c7631f0ab0f0db67248e216591048", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d53e1afaec330d21a2c7e42ba68fd696737bd38ba4f4d9067bd4b62aaa8f92a0", "generation_seconds_batch": 21.276315052993596, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.608595848083496, "target_z": -0.06738284310891599, "ce": 6.133810997009277, "genuineness": 4.118232727050781, "identity": 0.601752758026123, "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_CART_reg/046a90d0259f0270802273d9e45f1166ae6c7631f0ab0f0db67248e216591048.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "0501b83fe0bcbaa75586529f0f1ec389446e65a59e9e61aad5c14b2b5ef70174", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f3ee008090e3e2e2ecc6df6255555527a9a6b46dfe0a8e2f89a6018fb1be37fd", "generation_seconds_batch": 20.5459138690494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8171838521957397, "target_z": 0.08496090417029149, "ce": 6.09409761428833, "genuineness": 4.287685871124268, "identity": 0.6162122488021851, "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_CART_reg/0501b83fe0bcbaa75586529f0f1ec389446e65a59e9e61aad5c14b2b5ef70174.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "050600b43c0359c5418424b5a1fcce3f7fe13e41e5e3ff1e38cc528ff6eaa2fc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1f36315945259fb5ffee0adec1f84616b11845bcc424c1a76dcb01edba49febd", "generation_seconds_batch": 19.910247559077106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6279839277267456, "target_z": -0.05322261985939989, "ce": 6.173524379730225, "genuineness": 4.099404335021973, "identity": 0.5578639507293701, "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_CART_reg/050600b43c0359c5418424b5a1fcce3f7fe13e41e5e3ff1e38cc528ff6eaa2fc.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "052c5c9153c7f96bd45befa0025b8a63a6ceb725793e834a6488a25d1f74a154", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f308d07b8b7d9679e7a89c47d6ba12dff5f0e6d7a10bcd70104d0b1f3b2877d1", "generation_seconds_batch": 16.36496934306342, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9841689467430115, "target_z": -0.5234375163247552, "ce": 6.443574905395508, "genuineness": 4.13078498840332, "identity": 0.4252318739891052, "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_CART_reg/052c5c9153c7f96bd45befa0025b8a63a6ceb725793e834a6488a25d1f74a154.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": "056f5a6650f78e1c10109e27cb65168c3ded0ab35a5ec66b1fefd5690555b658", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e176684a1fa7de0101a027756880a473ed95aeb6a177f55c013e69c5dd789a82", "generation_seconds_batch": 15.236427792930044, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9895173907279968, "target_z": -0.5195312418376224, "ce": 6.157639026641846, "genuineness": 4.582659721374512, "identity": 0.5685758590698242, "hypothesis": "De sleutels liggen op de keukentafel. Ik kijk in de 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_CART_reg/056f5a6650f78e1c10109e27cb65168c3ded0ab35a5ec66b1fefd5690555b658.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "05813a5636f089aab5e19959a7e34a932a88b32acfd388fa7f51231edb16f708", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ce9e6eed426966991258b6614ac620b3dd1ab41ae398b4a60c80993e72715d99", "generation_seconds_batch": 18.895814412040636, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7469860315322876, "target_z": 0.03369142155445799, "ce": 5.792276382446289, "genuineness": 3.9676074981689453, "identity": 0.5024286508560181, "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_CART_reg/05813a5636f089aab5e19959a7e34a932a88b32acfd388fa7f51231edb16f708.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "c0d0b9c418aa4bf88a9ec93088eb6f053d6ab56c447b0244da12021f45330f88", "generation_seconds_batch": 15.435017101000994, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.39584386348724365, "target_z": -0.9531249673504896, "ce": 6.38797664642334, "genuineness": 4.375550270080566, "identity": 0.613104522228241, "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_CART_reg/059867605942d1f4c46c2c44d936e08b7fbe89dd4feef4c05500a7ed426c27a3.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "05d605701f14614970caa4c44ba59205957be8b9dfb7994de01cb0601d0a58cf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "dd175f0458537ec298c8649a2538ddb2763dbfd0988a68fb393af82377b7d47a", "generation_seconds_batch": 19.803269280004315, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7770707607269287, "target_z": 0.055664019647517625, "ce": 6.109982967376709, "genuineness": 4.143336772918701, "identity": 0.5159808397293091, "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_CART_reg/05d605701f14614970caa4c44ba59205957be8b9dfb7994de01cb0601d0a58cf.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "05eb8a6a42cc3a5a646d606fadf6ec6be696bdec01892263a14d0287e93507f6", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d1596d2cf36b683aaf0e3f8b127b93352778a2914703240609aa9cfa0213ed89", "generation_seconds_batch": 22.501690721022896, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.791931629180908, "target_z": 0.7968749455841494, "ce": 5.831989288330078, "genuineness": 3.9801597595214844, "identity": 0.47530248761177063, "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_CART_reg/05eb8a6a42cc3a5a646d606fadf6ec6be696bdec01892263a14d0287e93507f6.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "0613a1c5b42e60c488b2091db62e3aa21d0a43fb079a41330319554316cd4648", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "205f21970800f98d25f604dde20cc3b59e0df7e29e8a4bc420246b29bd2be853", "generation_seconds_batch": 20.433072619955055, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.374755620956421, "target_z": 0.4921874510257344, "ce": 5.879645347595215, "genuineness": 4.15588903427124, "identity": 0.6901389360427856, "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_CART_reg/0613a1c5b42e60c488b2091db62e3aa21d0a43fb079a41330319554316cd4648.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "061798a2dbe2e70198cd948215a8ea484de02e04cfd688a2ad524fb29d74d4fd", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "635f8fd3503988340c740514a766ec81677a8994d9b3e95ce9a85033e21d5feb", "generation_seconds_batch": 16.514050760073587, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.834718942642212, "target_z": 0.8281249673504896, "ce": 5.732706069946289, "genuineness": 3.7291181087493896, "identity": 0.34404098987579346, "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_CART_reg/061798a2dbe2e70198cd948215a8ea484de02e04cfd688a2ad524fb29d74d4fd.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "06477a1ebc6be541fef03ad9f7b740599f02416dd0c0736a3281e73823f26279", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d7f8da669f156f86a4e5f5ede92151fc12ad8db9593103c0ae78cfc7a7ae7847", "generation_seconds_batch": 21.751784158986993, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5459048748016357, "target_z": 0.6171875380910954, "ce": 5.792276382446289, "genuineness": 3.773050308227539, "identity": 0.4274609684944153, "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_CART_reg/06477a1ebc6be541fef03ad9f7b740599f02416dd0c0736a3281e73823f26279.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": "06db077ed59a457e5e8f71acaab2a1c13e45f99c3ac6a40f9b96f33d06c9ed79", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5191cc14864480b31b1b2094921a0edc7903bfbdb028873ec3888c1cc0649109", "generation_seconds_batch": 20.76387565396726, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9841689467430115, "target_z": -0.5234375163247552, "ce": 6.078212261199951, "genuineness": 4.300237655639648, "identity": 0.7746363878250122, "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_CART_reg/06db077ed59a457e5e8f71acaab2a1c13e45f99c3ac6a40f9b96f33d06c9ed79.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "070ae7ab6533021c189ca5308c3a62461e243d8946b04c55bcce4128032d2ab1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "df1bb4db52b31d57d5bc75bd12f76fbd1688c516ce0f7bb734fdf48a49a5cc7d", "generation_seconds_batch": 20.128580239019357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4817237854003906, "target_z": 0.570312418376224, "ce": 5.649308204650879, "genuineness": 3.9676074981689453, "identity": 0.3934592306613922, "hypothesis": "Encontrarlo la cajita. Estaba junto a la ventana, vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.11764705882352941, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/070ae7ab6533021c189ca5308c3a62461e243d8946b04c55bcce4128032d2ab1.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "071bc4fdb5e0d401df2c323ec037ffa47231eb17a7e099c0ee28a5d529bbc6c8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "edb4419804ea632e412a28941441150dfe7c29ca7546def733242392c2731ecc", "generation_seconds_batch": 15.658802633057348, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9896701574325562, "target_z": 0.2109375163247552, "ce": 6.173524379730225, "genuineness": 4.118232727050781, "identity": 0.36905497312545776, "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_CART_reg/071bc4fdb5e0d401df2c323ec037ffa47231eb17a7e099c0ee28a5d529bbc6c8.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "0734f27c34ebc4c701a5612dfa62164ed47d1d13af03fba87ce399b9af46290b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 129, "seconds_generated": 10.32, "mp3_sha256": "3f53b18d16f21fd75307918ada862909cfecb83ba75ae39e52e54b860eedd786", "generation_seconds_batch": 19.87202376802452, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.759841203689575, "target_z": 0.7734374727920746, "ce": 5.593709468841553, "genuineness": 3.961331605911255, "identity": 0.7057470083236694, "hypothesis": "I found the little box! It was beside the old window! We can look inside together!", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/0734f27c34ebc4c701a5612dfa62164ed47d1d13af03fba87ce399b9af46290b.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "07724fb7dce630b326ada297ed9716d6c61a90335eb370af97e8d6631448f89d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "13196ada1c970fbbaa3f67cf5f525d0739b59021f41812438d7f046c26a6ebfa", "generation_seconds_batch": 25.19867157097906, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.567298412322998, "target_z": 0.6328124619089046, "ce": 5.824047088623047, "genuineness": 3.8169827461242676, "identity": 0.3890873193740845, "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_CART_reg/07724fb7dce630b326ada297ed9716d6c61a90335eb370af97e8d6631448f89d.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "080463c53bef3f6bf9dbb3d619b3ae7c2a2a319947b0f7659a7c03c8116221a7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ed65a4f143f578f9da7e22d03a76165b2a25833815071606253a8a6a858c123e", "generation_seconds_batch": 21.702343598008156, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2411980628967285, "target_z": 1.124999912934639, "ce": 5.589737892150879, "genuineness": 3.948779582977295, "identity": 0.09252519905567169, "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_CART_reg/080463c53bef3f6bf9dbb3d619b3ae7c2a2a319947b0f7659a7c03c8116221a7.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": "0898b7aa98b67124f4d2179bd52ec6e37e195d4099fb210bec353837cae586b6", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f56beab33f46df264b495e699b09fcef2b9ff87e2f5ac47f220f0dc0c1aabd3d", "generation_seconds_batch": 20.35923095198814, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.235544204711914, "target_z": -0.339843790811888, "ce": 6.213237762451172, "genuineness": 4.444586753845215, "identity": 0.669471263885498, "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_CART_reg/0898b7aa98b67124f4d2179bd52ec6e37e195d4099fb210bec353837cae586b6.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "089d12322199580ec8399221023583fc45119a722f4de6d0daeb6a772083c8dc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "53cb82be8283f378bfdb5bcf2323180105fee7d22bbcb06e935433aeb49500ec", "generation_seconds_batch": 25.507614828064106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.134077310562134, "target_z": 0.3164063180198133, "ce": 6.221180438995361, "genuineness": 4.105680465698242, "identity": 0.0815172791481018, "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_CART_reg/089d12322199580ec8399221023583fc45119a722f4de6d0daeb6a772083c8dc.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "090c1e6c4c320a3778fad69e4cc9a43f9becd53b2082bf3c29c596922a62fe30", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "67226009b12cbe10c0c715bee1cdcfcdf001ec74883e11dbf5f9d887eef008e6", "generation_seconds_batch": 22.73456231795717, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.134077310562134, "target_z": 0.3164063180198133, "ce": 5.959072113037109, "genuineness": 3.9801597595214844, "identity": 0.07775239646434784, "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_CART_reg/090c1e6c4c320a3778fad69e4cc9a43f9becd53b2082bf3c29c596922a62fe30.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": "095bf72c4bc2d54ee0714fa0f6bbfc8e1e6bcc7d932cb528afe8f5c999268c84", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c04cd74778ac221c671486ac77a1d1f34e1cf75e62ce16262f49eab6d350e259", "generation_seconds_batch": 20.536085210973397, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.235544204711914, "target_z": -0.339843790811888, "ce": 6.197352409362793, "genuineness": 4.432034492492676, "identity": 0.3945510983467102, "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_CART_reg/095bf72c4bc2d54ee0714fa0f6bbfc8e1e6bcc7d932cb528afe8f5c999268c84.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "09cc399cd0c6d7e40c200a5e6a785c39abee1b0c2d8b2a39475212e4d0594ec7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3a75f635c2e89b693e0443feac6a06d59fceae0070acb7657ef0417b014a29ea", "generation_seconds_batch": 18.882992491940968, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5459048748016357, "target_z": 0.6171875380910954, "ce": 5.601652145385742, "genuineness": 3.860914945602417, "identity": 0.423048734664917, "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_CART_reg/09cc399cd0c6d7e40c200a5e6a785c39abee1b0c2d8b2a39475212e4d0594ec7.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": "09fd631cb75fc4ec2b0028ac22fc5492b0df7b770f31524acb6a09e8c4a007b1", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The keys are on the kitchen table.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2669be04a809aa98e1a745303a059d667822e7815a82fd6762420d55f1ea7f11", "generation_seconds_batch": 15.7713026299607, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1660149097442627, "target_z": -0.39062501088317014, "ce": 6.340320587158203, "genuineness": 4.030367851257324, "identity": 0.41408026218414307, "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_CART_reg/09fd631cb75fc4ec2b0028ac22fc5492b0df7b770f31524acb6a09e8c4a007b1.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "09ffcd135ce8fb83f09c2a62f057f56f6522c1815400f73dd2b09b53733d2ab4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "113c2b053373826bb0c41bc1dc09b8d2e4f603ae604fd867cd9fcb67c370766a", "generation_seconds_batch": 20.89842421002686, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5016276836395264, "target_z": -0.1455078104594056, "ce": 6.1020402908325195, "genuineness": 4.18099308013916, "identity": 0.6499620079994202, "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_CART_reg/09ffcd135ce8fb83f09c2a62f057f56f6522c1815400f73dd2b09b53733d2ab4.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "0aaf312863bb26644b0367ee0de6d0055ac9b1f3fc7710b1608f76dd2773816e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "259838acd5828d7bf7ac4bb0e62d795f4c49781318249133ef24f656b9a3574e", "generation_seconds_batch": 14.724796320078894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2301958799362183, "target_z": -0.34374997823365977, "ce": 6.324435234069824, "genuineness": 4.111956596374512, "identity": 0.5610706806182861, "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_CART_reg/0aaf312863bb26644b0367ee0de6d0055ac9b1f3fc7710b1608f76dd2773816e.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "0ad59ce30849720c2c667c09068e91dffdc149d779d1ef3a97ed2e776256f318", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "6fd2183e33f19e8d1f9878ae599d8dd8a6dc8663a2c08e7a13290affb01a6560", "generation_seconds_batch": 14.924323737970553, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6032475233078003, "target_z": -0.07128903053068775, "ce": 6.1894097328186035, "genuineness": 4.406930446624756, "identity": 0.593734860420227, "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_CART_reg/0ad59ce30849720c2c667c09068e91dffdc149d779d1ef3a97ed2e776256f318.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "0b0751f125802f628a006d564daac0598b76c062afdd329d46c10114302b5631", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "06477b9f12e56f87ebf9a5dba7d8fe2f5b8af6ac2dcadccc6cfb382a4cd329d9", "generation_seconds_batch": 16.089923155959696, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8558071255683899, "target_z": -0.6171874945584149, "ce": 6.260893821716309, "genuineness": 4.507347106933594, "identity": 0.48101842403411865, "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_CART_reg/0b0751f125802f628a006d564daac0598b76c062afdd329d46c10114302b5631.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "0b621c53c5da337467898398cb549d2eb746154691a4e33a1fc3bb37a235e2a8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ead27a762ba870ccf52fe7a1eb7ffc6f2b8a9583871914d235a66e776feeeebf", "generation_seconds_batch": 20.10162755602505, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.814509630203247, "target_z": 0.0830077669267251, "ce": 6.205295085906982, "genuineness": 4.218649387359619, "identity": 0.5651824474334717, "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_CART_reg/0b621c53c5da337467898398cb549d2eb746154691a4e33a1fc3bb37a235e2a8.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "0be808322896046c39ef3240022cb695094e63fec47eb76a350b6da6e78da708", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3f7f786ba9db6c4b65cb17f9510840f851a5334ec32661500039f2e33da0da95", "generation_seconds_batch": 22.546282201074064, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.695660352706909, "target_z": 0.7265625272079254, "ce": 5.831989288330078, "genuineness": 4.005263805389404, "identity": 0.41421517729759216, "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_CART_reg/0be808322896046c39ef3240022cb695094e63fec47eb76a350b6da6e78da708.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "974e512a46e80f36dec80a472bafc1f97fa314376853a361b93716332d3feab4", "generation_seconds_batch": 14.61928462004289, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.24608826637268066, "target_z": -1.0625000435326806, "ce": 6.197352409362793, "genuineness": 4.2249250411987305, "identity": 0.596225380897522, "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_CART_reg/0cb70e86df8754d3bba44fe7f798aac7d0ffe2ca68fc75b92ff1a1ed29986ba7.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "0d14fdfc9dbf04484e20edf353dd84efa1321348ececb97bd659671646ba87a1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3ed2c70fa1ed52b48c5737e235aaf24a2fb9f254067dded901cb8109aa5ac52f", "generation_seconds_batch": 23.051225333008915, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.856112480163574, "target_z": 0.8437498911682987, "ce": 5.752562999725342, "genuineness": 4.005263805389404, "identity": 0.5863142013549805, "hypothesis": "¡Eh! He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/0d14fdfc9dbf04484e20edf353dd84efa1321348ececb97bd659671646ba87a1.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "0df7d32adada75cfb6373fea515ff58d9c6291f71f34324aecbfb67f48e0c980", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ec3b09808e56887946235ec1772c9290f56e35411bc486fbd4d2949dc886df17", "generation_seconds_batch": 20.200493552954867, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4121944904327393, "target_z": 0.5195311983049419, "ce": 6.070269584655762, "genuineness": 3.9299514293670654, "identity": 0.5560674667358398, "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_CART_reg/0df7d32adada75cfb6373fea515ff58d9c6291f71f34324aecbfb67f48e0c980.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "0e09a5e4e5a670bbbc3cdc0fd63e1a853d9fc66f2e36a4dfa698ab680b810fb5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "65fb2f00d67c56f194661122cdad9d2a229d1916de5e3d7cd80f12d62c04883a", "generation_seconds_batch": 21.712897877092473, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0511767864227295, "target_z": 0.2558593246653382, "ce": 5.895530700683594, "genuineness": 4.262581825256348, "identity": 0.6390341520309448, "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_CART_reg/0e09a5e4e5a670bbbc3cdc0fd63e1a853d9fc66f2e36a4dfa698ab680b810fb5.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "0eff81f674c500281e8fd53000ce0746e0d1111e1da3037aa4d236ab66e277e8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4154af44d7a4ef882979a0f6d0f5e455f362a44efdd1c5e4dcc3a34d03730190", "generation_seconds_batch": 15.708798031904735, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.917466640472412, "target_z": 0.15820315900990664, "ce": 5.708878040313721, "genuineness": 4.18099308013916, "identity": 0.4300149083137512, "hypothesis": "La porte t'a encore ouverte ! J'ai posé ton manteau sur la chaise. Dis-moi, quand toi y est prêt ?", "wer": 0.2222222222222222, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/0eff81f674c500281e8fd53000ce0746e0d1111e1da3037aa4d236ab66e277e8.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "0f4fb26ce6812591c9ff19cb08b605c5d2f81b79f8456f17fb1f2e4b4359e3d6", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e88c6f7c62381d4186abe6c7b950e2a9a5ca73891a813c9dd2c3c0b236e28866", "generation_seconds_batch": 18.87684214406181, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7797449827194214, "target_z": 0.05761715689108402, "ce": 6.006728172302246, "genuineness": 3.9299514293670654, "identity": 0.5088661909103394, "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_CART_reg/0f4fb26ce6812591c9ff19cb08b605c5d2f81b79f8456f17fb1f2e4b4359e3d6.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": "0f7c602908f5bce02d1b93d8179c3e74c98030da42d4ab11f5f4f1d2755cc308", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "762f3f3e9bb9cfa14b27c65b15bb0225572cbdc40bd9c4b274069fbb30e21721", "generation_seconds_batch": 20.75274171505589, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3318157196044922, "target_z": -0.2695311983049419, "ce": 6.2767791748046875, "genuineness": 4.256305694580078, "identity": 0.6878929138183594, "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_CART_reg/0f7c602908f5bce02d1b93d8179c3e74c98030da42d4ab11f5f4f1d2755cc308.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "0fac6d5efd604c9cf24d48fd1d605f7c20840e65ad3689e0d6722f38611a9d36", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "6799a3567da34c33b4f1110535f0079371128ef936f9da320c6da0da13bfdfb0", "generation_seconds_batch": 25.81117937306408, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6528730392456055, "target_z": 0.6953125054415851, "ce": 5.903473377227783, "genuineness": 4.036643981933594, "identity": -0.027755986899137497, "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_CART_reg/0fac6d5efd604c9cf24d48fd1d605f7c20840e65ad3689e0d6722f38611a9d36.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "0fce6d232793518316201156d541e98d0d652b7d8c12f59b0902b3e6aee36cae", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a77f0c13a62e34a792a46c91646fe28564b04019b11cc17552ea54947f687f76", "generation_seconds_batch": 20.769450145075098, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7170538902282715, "target_z": 0.7421874510257345, "ce": 5.895530700683594, "genuineness": 3.8044304847717285, "identity": 0.2552269697189331, "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_CART_reg/0fce6d232793518316201156d541e98d0d652b7d8c12f59b0902b3e6aee36cae.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "0fe8d88ede90564e4601c1d33b6236294125465cd4301c460ed63bdc4afff18f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "74b68314c52f1ba6bc00cb1f1ae56b86f32a442a643e1c1aeeb0c3eead33f73b", "generation_seconds_batch": 19.168484138092026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.638012170791626, "target_z": -0.04589842049504668, "ce": 6.173524379730225, "genuineness": 4.356721878051758, "identity": 0.32715365290641785, "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_CART_reg/0fe8d88ede90564e4601c1d33b6236294125465cd4301c460ed63bdc4afff18f.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "106aa8eb115b5f1a67f36208c477d38e650ca058b75a26660037da41c0a41597", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "68533bc4cbf21858bcc8258cea819b17e456cfe1194322f75b877955aff66dba", "generation_seconds_batch": 21.610666926018894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2223260402679443, "target_z": 0.38085941173069915, "ce": 6.006728172302246, "genuineness": 3.9550557136535645, "identity": 0.504383385181427, "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_CART_reg/106aa8eb115b5f1a67f36208c477d38e650ca058b75a26660037da41c0a41597.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": "1071cdfa6b261535793694092f621f519e329a9088aff8fd0a224c9f8dc9faab", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "31b758782290d4386181e0002fcab68c2831a221fdb00550f846a62bb2a381ce", "generation_seconds_batch": 19.73768907203339, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.968123733997345, "target_z": -0.5351562527207925, "ce": 5.919358730316162, "genuineness": 4.118232727050781, "identity": 0.6383328437805176, "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_CART_reg/1071cdfa6b261535793694092f621f519e329a9088aff8fd0a224c9f8dc9faab.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "1106c0624bdb86b011efcdbd41876572a456b0c5ba02ca2ef6a3efa57882ba28", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d7932a1d91089f03e8db0b8d145fd11c39e6c1d30773ea7883a7e1f3a6e7376e", "generation_seconds_batch": 15.725315557909198, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2836799621582031, "target_z": -0.30468749455841493, "ce": 6.165581703186035, "genuineness": 3.986435890197754, "identity": 0.4472874104976654, "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_CART_reg/1106c0624bdb86b011efcdbd41876572a456b0c5ba02ca2ef6a3efa57882ba28.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "114189c886aff43755e892dfc2b8ae3ba90b5a1fedd2b6af962add82b3075a60", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "94b448798f745e12f288a826410320f5aa191d16afd245cb6f482eee3f16fd5a", "generation_seconds_batch": 21.322745061013848, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0486555099487305, "target_z": 0.9843750761821909, "ce": 5.6254801750183105, "genuineness": 3.8797430992126465, "identity": 0.3601524233818054, "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_CART_reg/114189c886aff43755e892dfc2b8ae3ba90b5a1fedd2b6af962add82b3075a60.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "117e2a7ba2efb979fab6286f26fc79ffe45ac4b74de5a7aa8469a10a6a069e6c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bd7673865770a16936a602de6d97568ae036fab5fb0c97386e1f6ed0431425ad", "generation_seconds_batch": 21.751784158986993, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.26259183883667, "target_z": 1.1406250108831701, "ce": 5.530168056488037, "genuineness": 3.6851859092712402, "identity": 0.4609471261501312, "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_CART_reg/117e2a7ba2efb979fab6286f26fc79ffe45ac4b74de5a7aa8469a10a6a069e6c.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "11db60b7a3034cf0f29193829412839d0cc1020f95c4369a53e03dab008d4ea9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ae7ebf3df0219383cdad9291c593968f0242dd83d49b1e0f46d08dcda9bee4e8", "generation_seconds_batch": 20.5459138690494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5484262704849243, "target_z": -0.11132812636039627, "ce": 6.046441555023193, "genuineness": 4.306513786315918, "identity": 0.4741377532482147, "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_CART_reg/11db60b7a3034cf0f29193829412839d0cc1020f95c4369a53e03dab008d4ea9.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": "1207a8d27a6578b53790fe8f9638b256a4fd088955f51d82a44374a6d19e2fd8", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f98f8a2429ea028bf4c4e93eba5406630611ca74be5241956aec13af5d3e3f62", "generation_seconds_batch": 19.841244725044817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.722096860408783, "target_z": -0.7148437472792075, "ce": 6.181467056274414, "genuineness": 4.293961524963379, "identity": 0.6680659055709839, "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_CART_reg/1207a8d27a6578b53790fe8f9638b256a4fd088955f51d82a44374a6d19e2fd8.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "1250e57794afd273ed8604090d5b3f722b4d4bd664668f6d678dd887abbaf95b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a122204a18dabe95af3fb188e5b10fa59e6cf70a3e73fa6418343204f754a904", "generation_seconds_batch": 21.335506377043203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4107047319412231, "target_z": -0.2119140414138579, "ce": 6.300607204437256, "genuineness": 4.15588903427124, "identity": 0.435099333524704, "hypothesis": "Ich habe die kleine Schachtel gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschaffen.", "wer": 0.0625, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/1250e57794afd273ed8604090d5b3f722b4d4bd664668f6d678dd887abbaf95b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "12648b07af46f045929300765de0dcf32e8e01b218b72b545cf82811f8b37570", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "476e254f3ea97cf45220a4d35e4bb000c584961e9c44a1d49581d5960b291918", "generation_seconds_batch": 20.504617818049155, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5459048748016357, "target_z": 0.6171875380910954, "ce": 6.046441555023193, "genuineness": 4.281409740447998, "identity": 0.20518279075622559, "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_CART_reg/12648b07af46f045929300765de0dcf32e8e01b218b72b545cf82811f8b37570.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": "12850f08cd6b14974307143b4ca959c6716319e8fdd8e54abf94c698c4cf6f36", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "02ad1fea746f74162cebaf935c073ba6650633c81ba97682443d2193cb7a9816", "generation_seconds_batch": 16.045790812000632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9360332489013672, "target_z": -0.5585937690455477, "ce": 6.23706579208374, "genuineness": 4.080576419830322, "identity": 0.5838794112205505, "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_CART_reg/12850f08cd6b14974307143b4ca959c6716319e8fdd8e54abf94c698c4cf6f36.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "1296710fd2e88af754da6ce8dc62570db7286392f483c4c55d3a140c29d1d4d4", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f150dc8324d485c4f0a8d5258840e8c13422a07cd848f6cb7fd499183094e49f", "generation_seconds_batch": 15.419269283069298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6739611625671387, "target_z": -0.75, "ce": 6.41180419921875, "genuineness": 4.162164688110352, "identity": 0.5530047416687012, "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_CART_reg/1296710fd2e88af754da6ce8dc62570db7286392f483c4c55d3a140c29d1d4d4.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "2e1441d21359947b4c2543bcce1abb0cd6a7b88d6a7181f1c9729dcd2527a695", "generation_seconds_batch": 14.715727481991053, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.36910176277160645, "target_z": -0.9726562527207925, "ce": 6.1894097328186035, "genuineness": 4.231201171875, "identity": 0.6496502161026001, "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_CART_reg/12a7b850c68b1e935d69d2b179f3c0df54c4775061935760902fd9f6cbfc7541.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "12e2fcd6c8ceb7f16169bc568d74eb9244bf30848c9945d5c69264b886d61f7f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fa5d0e910b32e8bc355d2e638cbb70bbf4438f9ce7f86b753a11fe9f71377b57", "generation_seconds_batch": 20.206531993928365, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0536983013153076, "target_z": -0.47265625272079254, "ce": 6.173524379730225, "genuineness": 4.457139015197754, "identity": 0.4522995352745056, "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_CART_reg/12e2fcd6c8ceb7f16169bc568d74eb9244bf30848c9945d5c69264b886d61f7f.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "130f344dc676a25f593f18a08bb531b87b8c5f91dbcc744f7f9ccadd5c1b1ce2", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "31f2a940b9bca5d307c7fecc46f6892eabe96227aa014a8d751d3fcc5a283653", "generation_seconds_batch": 15.284291871939786, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.925336480140686, "target_z": -0.5664062309544523, "ce": 6.292664527893066, "genuineness": 4.482243061065674, "identity": 0.5843448638916016, "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_CART_reg/130f344dc676a25f593f18a08bb531b87b8c5f91dbcc744f7f9ccadd5c1b1ce2.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "139fa5ee151c58d43fa7e4e638579fc94f8d557c97fbad336e162177a0fecbc2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4beaa0dffedf6a6eda779c51c68fd92491a4d623237421cf691ee5ee811e5379", "generation_seconds_batch": 22.17708530300297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6620800495147705, "target_z": -0.028320272368310157, "ce": 5.816103935241699, "genuineness": 4.369274139404297, "identity": 0.4874785542488098, "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_CART_reg/139fa5ee151c58d43fa7e4e638579fc94f8d557c97fbad336e162177a0fecbc2.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "14270ac9475de483f1298acecda3f19cd9c1f4b66beaaf7ba02291193acb492e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "cdbfa224d8218ef6f6f4c15e0e849395c468efc4a8de25cc7a19276ae04015dc", "generation_seconds_batch": 15.880435994942673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4603302478790283, "target_z": 0.5546874945584149, "ce": 5.454712867736816, "genuineness": 3.8546390533447266, "identity": 0.6573668718338013, "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_CART_reg/14270ac9475de483f1298acecda3f19cd9c1f4b66beaaf7ba02291193acb492e.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "15037db0cb765ae5f3c8d47e610fff0fa2f416106b6e2e27c3ea38231e9a8547", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "593db55ff31b4274e4059cdd390d3bfadd8ba5502143f0fc1bc00cb24b93cc1d", "generation_seconds_batch": 14.924323737970553, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6674283742904663, "target_z": -0.024414084946538385, "ce": 6.0146708488464355, "genuineness": 4.394378662109375, "identity": 0.6951707601547241, "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_CART_reg/15037db0cb765ae5f3c8d47e610fff0fa2f416106b6e2e27c3ea38231e9a8547.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": "15257d6e3d96df84bf03ef8ea2b45d67e1c2148bac1cbd7d462274a9b03abe6b", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "86660e57a19784ee507189fa66d4cd5d32e09fa142a53c5dde785e2f6744e354", "generation_seconds_batch": 20.08546127693262, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2703089714050293, "target_z": -0.31445309371088587, "ce": 6.109982967376709, "genuineness": 4.243753433227539, "identity": 0.5871737003326416, "hypothesis": "De weg gaat linksaf naar de brug. Nop 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_CART_reg/15257d6e3d96df84bf03ef8ea2b45d67e1c2148bac1cbd7d462274a9b03abe6b.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "15413c11cdfcd6044643396155872abd03fb38bd5ff65844225b7eaa196a3f42", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f3db4025fce79aa25678e9ae1b07e9617f5b1d8a32fac7f22d589c54f0852808", "generation_seconds_batch": 19.91401848499663, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8639824390411377, "target_z": 0.11914058826930082, "ce": 6.070269584655762, "genuineness": 4.331617832183838, "identity": 0.4997548460960388, "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_CART_reg/15413c11cdfcd6044643396155872abd03fb38bd5ff65844225b7eaa196a3f42.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "15475a46c9ab75e847d49e03d970375a459114bb921f78a2391635fe5e46bc4b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7cfa69260917425522e9d502239125806c2f4cc622b7e2b96302558cecdb0739", "generation_seconds_batch": 26.60732024302706, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.738447666168213, "target_z": 0.7578125489742655, "ce": 5.422942161560059, "genuineness": 3.7793264389038086, "identity": -0.05586787685751915, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude ruim. 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_CART_reg/15475a46c9ab75e847d49e03d970375a459114bb921f78a2391635fe5e46bc4b.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "2f05d81368da826b35a0dd4cf3ab6d3cdbf1a69154cc096fba05e32d04a0be74", "generation_seconds_batch": 15.132483097026125, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.2353914976119995, "target_z": -1.070312505441585, "ce": 6.300607204437256, "genuineness": 4.469690799713135, "identity": 0.7341668605804443, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zum kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/15585f3962d31fd159bb39a755593ab24aa0df87e4725d7772f40811258760f5.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "1643e4a21794db75a4c1f44c83dff957f5885f067085fdd1fa8c4d263b64f2fc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b57c83fe560038bb4d1f8d116e9bba9f135d5c33b5a6e2adcbda654fbc56c038", "generation_seconds_batch": 15.105503295082599, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1634933948516846, "target_z": 0.33789056650296057, "ce": 5.863759994506836, "genuineness": 3.860914945602417, "identity": 0.3343827724456787, "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_CART_reg/1643e4a21794db75a4c1f44c83dff957f5885f067085fdd1fa8c4d263b64f2fc.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": "16aaf7f9e52b42317a4d1402fa4e83bc612dd6c08be226110544a19b7c0198e9", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c846f6b781615b52997699b9fd915513b0f72dd2ad16b1e912ba94f4228500c9", "generation_seconds_batch": 14.940477830008604, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0002142190933228, "target_z": -0.5117187363960374, "ce": 5.879645347595215, "genuineness": 4.457139015197754, "identity": 0.7103524804115295, "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_CART_reg/16aaf7f9e52b42317a4d1402fa4e83bc612dd6c08be226110544a19b7c0198e9.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "16b70f2d00d6ad62a853e57c06a6ece6a59f229935bcf241f38dc8fc59c7a753", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "752bbd33ff00d43d3f109656481e6b2f45c44aece551a375fefcc2d71b9ac269", "generation_seconds_batch": 20.231323575950228, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9948657751083374, "target_z": -0.5156250108831701, "ce": 6.125868320465088, "genuineness": 4.256305694580078, "identity": 0.7387826442718506, "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_CART_reg/16b70f2d00d6ad62a853e57c06a6ece6a59f229935bcf241f38dc8fc59c7a753.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "16b8e402789891dd80425d5ae2eab5ed971f114a44bb88d9a76acb44efcd7adc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "6e9dd61c8d5feb995fbd22bab7d5243eebbead7cb9eb70872f804fd3b66ba405", "generation_seconds_batch": 21.335506377043203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6968446969985962, "target_z": -0.002929662332669083, "ce": 6.3641486167907715, "genuineness": 4.105680465698242, "identity": 0.29932790994644165, "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_CART_reg/16b8e402789891dd80425d5ae2eab5ed971f114a44bb88d9a76acb44efcd7adc.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "16d56ff83838c8f245305248b190a2cd8529a9ab9070fb9460315650ead48620", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "156ff1525a1d6a59f125089927da57026df82a46bea4b1c8229bb2e1b958a3e0", "generation_seconds_batch": 20.24757582996972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8933987617492676, "target_z": 0.14062501088317012, "ce": 6.078212261199951, "genuineness": 4.2374773025512695, "identity": 0.7657729387283325, "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_CART_reg/16d56ff83838c8f245305248b190a2cd8529a9ab9070fb9460315650ead48620.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": "1744cb8f608c50e4b9d0235dcbd98a774e8ca4a94f5564e700c7ea6a1786eae1", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "905219c67ae21dbab77eaf4c0efc908eb4f4f126fc58c83a5219a5c49f832de6", "generation_seconds_batch": 15.465044792043045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.599083423614502, "target_z": -0.8046874945584149, "ce": 6.2767791748046875, "genuineness": 4.6705241203308105, "identity": 0.574543833732605, "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_CART_reg/1744cb8f608c50e4b9d0235dcbd98a774e8ca4a94f5564e700c7ea6a1786eae1.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "175b2fdc2e298dbbfb6d66cac53be70549a2fa80a463cd934ac140e94cb9b684", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d9d7023372fee26bcd6f36679353c8ea633ace6e69f5dab9406dc73fc2ddf81e", "generation_seconds_batch": 15.658802633057348, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5940403938293457, "target_z": 0.6523436602138465, "ce": 6.149696350097656, "genuineness": 3.9173994064331055, "identity": 0.7179421782493591, "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_CART_reg/175b2fdc2e298dbbfb6d66cac53be70549a2fa80a463cd934ac140e94cb9b684.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "17b79841c9ccb4a9c0f982144e3a33c5b0b3c84aefce3077132c30251cf29314", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4c7d00d9b5a6582d3b27788e056c29034c5756f90d5c588433294922e989682e", "generation_seconds_batch": 20.161289564915933, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.187561273574829, "target_z": 0.3554687146296971, "ce": 5.7684478759765625, "genuineness": 4.105680465698242, "identity": 0.47185218334198, "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_CART_reg/17b79841c9ccb4a9c0f982144e3a33c5b0b3c84aefce3077132c30251cf29314.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": "18eb2626672adef2a26a631fd889799c26f66faee49ca254f698c9938d9ad2b5", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5e9101c78f091c45ff8f40284c5cace4ee0e44c98fbddce1f81312497699293e", "generation_seconds_batch": 20.417473209905438, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6846579313278198, "target_z": -0.7421875380910954, "ce": 6.197352409362793, "genuineness": 4.607763767242432, "identity": 0.5780966877937317, "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_CART_reg/18eb2626672adef2a26a631fd889799c26f66faee49ca254f698c9938d9ad2b5.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "193174198021009a652fa1d1f16d23471ac4088c7d2730d6f58512d1aff8a000", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "da33582ca85b9d141d2ac5004af4be3f1e946c396ceadd4d2021448edd6f225f", "generation_seconds_batch": 18.941351162968203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.930837631225586, "target_z": 0.1679687581623776, "ce": 5.927301406860352, "genuineness": 4.325342178344727, "identity": 0.02471441775560379, "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_CART_reg/193174198021009a652fa1d1f16d23471ac4088c7d2730d6f58512d1aff8a000.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "1977b8056a87c6ab3e584956599aab36866191559f1209fb42b05279e3ac745c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "03107b7568bbdcebc6fba085faa2fcf8aaf6448333aecd29f7c09c5b8ba763a3", "generation_seconds_batch": 19.58179482491687, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9841689467430115, "target_z": -0.5234375163247552, "ce": 6.443574905395508, "genuineness": 4.231201171875, "identity": 0.4016619324684143, "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_CART_reg/1977b8056a87c6ab3e584956599aab36866191559f1209fb42b05279e3ac745c.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "19af89a6ac3722f0d3f054e582e9160c8deea7ae9337b393fc9247d1e41fe557", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "185c34ab221911aae85add44a745e4d35636d0f1e68a294bcbdbfdfa6424a14b", "generation_seconds_batch": 21.191109837032855, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2918553352355957, "target_z": 0.4316406318019813, "ce": 5.990842819213867, "genuineness": 4.18726921081543, "identity": 0.4750175476074219, "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_CART_reg/19af89a6ac3722f0d3f054e582e9160c8deea7ae9337b393fc9247d1e41fe557.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": "19d948967aa5d4e4562d89db23c2f2e78e345367955724dcb94486a323d059cc", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9e589d16cc51f2f9441a8206f9b98edc518943f55f603265a2975134127ea646", "generation_seconds_batch": 16.439308190951124, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.39584386348724365, "target_z": -0.9531249673504896, "ce": 6.23706579208374, "genuineness": 4.607763767242432, "identity": 0.6658528447151184, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter. Bis zum kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/19d948967aa5d4e4562d89db23c2f2e78e345367955724dcb94486a323d059cc.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "1a4ee7d9f85835f116f963b82655a22954b4e2c307f9dac1fb0fdca0ab620721", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f586c94e5ee3bee4f90f56265f8420db8b62db273a363b6fa2d09ede1757dc9c", "generation_seconds_batch": 20.011779442080297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.056525230407715, "target_z": 0.25976559915247094, "ce": 6.0305562019348145, "genuineness": 4.20609712600708, "identity": 0.27540987730026245, "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_CART_reg/1a4ee7d9f85835f116f963b82655a22954b4e2c307f9dac1fb0fdca0ab620721.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "1b0708e721a46802ed2a24e2b4dc7453f9cd5e694a6b381021bc53eefe35a3db", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "dd382ee53e43098117d735408ffa1790a27d0f302270d11e18f23348853365bf", "generation_seconds_batch": 19.841903843916953, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.021913528442383, "target_z": 0.964843877877249, "ce": 5.375286102294922, "genuineness": 3.7416701316833496, "identity": 0.24575793743133545, "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_CART_reg/1b0708e721a46802ed2a24e2b4dc7453f9cd5e694a6b381021bc53eefe35a3db.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "1b7a19a0703f380da9c23e42645751efd213a09f72b38c650851ad9f40889de2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b1e4f344c08f797482e42092d023e13912723aec3802f18d52390505befd9beb", "generation_seconds_batch": 19.32955087092705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5323810577392578, "target_z": -0.1230468627564336, "ce": 6.308549880981445, "genuineness": 4.256305694580078, "identity": 0.41252830624580383, "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_CART_reg/1b7a19a0703f380da9c23e42645751efd213a09f72b38c650851ad9f40889de2.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "1b7b6af3edac1bed3b53b66d0eec9002826d54e3285ca1323ae61dff4a67593d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a757e86683c4078d8e1413b1da314c7784cc31c67cfd452b7fb05d2c81a837a9", "generation_seconds_batch": 21.628031324013136, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.2781788110733032, "target_z": -1.039062483675245, "ce": 6.403861999511719, "genuineness": 4.319066047668457, "identity": 0.5185990929603577, "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_CART_reg/1b7b6af3edac1bed3b53b66d0eec9002826d54e3285ca1323ae61dff4a67593d.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "1bd01685818781f3eca998f67bbc94d959c8143956c8a754a0779849d88743f4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0ad7960c7907bba710219da6c5fcc15002014b8254aaf8a24cfad20f1ed834d8", "generation_seconds_batch": 19.706848094007, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2223260402679443, "target_z": 0.38085941173069915, "ce": 5.998785495758057, "genuineness": 4.21237325668335, "identity": 0.5857386589050293, "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_CART_reg/1bd01685818781f3eca998f67bbc94d959c8143956c8a754a0779849d88743f4.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "1d0324c543ecf9933628c5416715a99cba07b5a150a71bf6ebcc2200a08cb5d3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "85acdc017bff23d8494d463fee2418c9a71ad1187087d68a350fb3def5726745", "generation_seconds_batch": 20.653390570892952, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.834413468837738, "target_z": -0.6328125054415851, "ce": 6.165581703186035, "genuineness": 4.356721878051758, "identity": 0.7106059193611145, "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_CART_reg/1d0324c543ecf9933628c5416715a99cba07b5a150a71bf6ebcc2200a08cb5d3.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "1d1790063ef5e372af1d69854fc580ede38259e22d8e9355ca504ec17cfe870c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3322885527b65c15f389c9bf15f7b6779adb0b918bc3b821ec62b00d001856d1", "generation_seconds_batch": 16.355668568983674, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4468064308166504, "target_z": -0.1855469062891141, "ce": 6.149696350097656, "genuineness": 4.080576419830322, "identity": 0.4721212089061737, "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_CART_reg/1d1790063ef5e372af1d69854fc580ede38259e22d8e9355ca504ec17cfe870c.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "1d17c8a692255420944d05e01c02cc9ee431d07f883653c54844ea7b5be1a5f0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5680dadc8e6ddb56e7998e2d49c3618518c48e4162e12dd4e02cab4918281f7d", "generation_seconds_batch": 21.071923605981283, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.956242561340332, "target_z": 0.1865234313782168, "ce": 6.0146708488464355, "genuineness": 4.243753433227539, "identity": 0.42369794845581055, "hypothesis": "In concordó la cajita. Estaba junto a la ventana vieja. ¿Podemos mirar juntos lo que hay dentro?", "wer": 0.11764705882352941, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/1d17c8a692255420944d05e01c02cc9ee431d07f883653c54844ea7b5be1a5f0.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "1d33a122a4bae26204b528ee25a80c4e1cb589a5dee916a626cb2b1ecd113e3f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "50e459bcc3d11ffa3f5c2951c7aca302f2453c3b5d869e3a322942eab7a8d017", "generation_seconds_batch": 21.80552641209215, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5459048748016357, "target_z": 0.6171875380910954, "ce": 6.125868320465088, "genuineness": 4.293961524963379, "identity": 0.5043003559112549, "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_CART_reg/1d33a122a4bae26204b528ee25a80c4e1cb589a5dee916a626cb2b1ecd113e3f.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "c94056037941ff0be84b92cef93b04f65a410b703b2b3f0f4440cb74dc827d46", "generation_seconds_batch": 15.885165423038416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7434905171394348, "target_z": -0.6992187363960374, "ce": 6.205295085906982, "genuineness": 4.293961524963379, "identity": 0.6122477650642395, "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_CART_reg/1d559e6da25f8bbacc3b53d5880369445633ce7ab0125ca7b07093108b8c8de3.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "1d60f6365f9b782a15b5b94764ee44b62f44caeceee176862758990d5455a741", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "80e9c61d7ce108cbf7236432852d1a7ff8a25ba24d0b8a39e70b0f591dc84fec", "generation_seconds_batch": 20.674201525049284, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3373167514801025, "target_z": 0.46484370374652695, "ce": 5.998785495758057, "genuineness": 3.9927117824554443, "identity": 0.18449103832244873, "hypothesis": "Ich habe die kleine Schacht gefunden. Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen.", "wer": 0.0625, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/1d60f6365f9b782a15b5b94764ee44b62f44caeceee176862758990d5455a741.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "0faac10149bd1657bd76453bad609d78056c15851d9c88ea084638952f412a0e", "generation_seconds_batch": 14.678812328027561, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.3744502067565918, "target_z": -0.9687499782336597, "ce": 6.41180419921875, "genuineness": 4.419482707977295, "identity": 0.5675771236419678, "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_CART_reg/1d829402ab2176d2024ae3e918286a39ec094d9a5299caa0e048ba37912e1579.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "1e31861d4b09830b5229fb12bad2d4925e15f6c73563dec71a56c4729a01271f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5e4310884a54f7b05616bfa664baafedb5714cdd44dbf4596ab1137c28d91140", "generation_seconds_batch": 15.883290670928545, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8880503177642822, "target_z": 0.13671873639603735, "ce": 5.895530700683594, "genuineness": 4.319066047668457, "identity": 0.5927572250366211, "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_CART_reg/1e31861d4b09830b5229fb12bad2d4925e15f6c73563dec71a56c4729a01271f.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "1e437223389adae6cf303b7cd976ec3774482f27f6536828969fa8a2202bb9db", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ff233890374a5aa379d97dbbaa799f46a78e047e64c208155628f9052054795e", "generation_seconds_batch": 21.274790597963147, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4628517627716064, "target_z": -0.17382808282771575, "ce": 6.1894097328186035, "genuineness": 4.162164688110352, "identity": 0.5207206010818481, "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_CART_reg/1e437223389adae6cf303b7cd976ec3774482f27f6536828969fa8a2202bb9db.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "1e89462a93624aefc654c144c25e47e56a1bbad6fa7510da8d301809c1f98313", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c8423b033802a377efa346c9c70e71669c45d33f0aff240328d6c02be32362c6", "generation_seconds_batch": 15.080126675078645, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6226354837417603, "target_z": -0.05712889434653268, "ce": 5.9431867599487305, "genuineness": 4.319066047668457, "identity": 0.6111276149749756, "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_CART_reg/1e89462a93624aefc654c144c25e47e56a1bbad6fa7510da8d301809c1f98313.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "3207fa32827cfc32972926b756dd74019059003c2b377703d60e5e53417283d6", "generation_seconds_batch": 15.885165423038416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7488389015197754, "target_z": -0.6953125054415851, "ce": 6.213237762451172, "genuineness": 4.2249250411987305, "identity": 0.636305570602417, "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_CART_reg/1eaf0e0e2c966186cf6ada2de41819a315f8b1b1eeeeef5a84a28dd582998db8.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "1ee29d063d6d937d2b5faf6490a7d6c641400592e007bb38c9136cf8b871dee4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c35cc51484b4d3380e0672fd574f41ae294956adcc6248f54dcc8b7e9c88240d", "generation_seconds_batch": 19.864074348006397, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3598947525024414, "target_z": -0.2490234749108973, "ce": 6.332377910614014, "genuineness": 4.3504462242126465, "identity": 0.35876309871673584, "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_CART_reg/1ee29d063d6d937d2b5faf6490a7d6c641400592e007bb38c9136cf8b871dee4.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": "1ee3596cafcaff35a3ac4ea0fde83bd3eec52fc51e2f41c1657c92f35c347bb9", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "83ebaedc46187c8c06c85c2e29323a44d8c86d1243567da59980041e05492cde", "generation_seconds_batch": 16.045790812000632, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3652431964874268, "target_z": -0.24511720042376453, "ce": 6.284721851348877, "genuineness": 4.055472373962402, "identity": 0.5103748440742493, "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_CART_reg/1ee3596cafcaff35a3ac4ea0fde83bd3eec52fc51e2f41c1657c92f35c347bb9.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "1f166702e8dfe19cc7be116546719c0d60c270c2f2d66f936de2f23549d8ddf0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a7231b080870241eca65c4c9b90b586b4efa81433ff292b6d67a05ac24e8bdad", "generation_seconds_batch": 22.19703052204568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5751683712005615, "target_z": -0.09179684099009336, "ce": 6.284721851348877, "genuineness": 4.444586753845215, "identity": 0.6241902112960815, "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_CART_reg/1f166702e8dfe19cc7be116546719c0d60c270c2f2d66f936de2f23549d8ddf0.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "1f1aca01b83efe0a634b85e806fc20a58b10a17ad992bdb62a29b7d07ee574f7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "6b81b4290224bf377ea6a820d47edeba560478da6db7f64a18da7ddd32697a52", "generation_seconds_batch": 20.63154619303532, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4013450145721436, "target_z": -0.21874997823365974, "ce": 6.2767791748046875, "genuineness": 4.607763767242432, "identity": 0.617351770401001, "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_CART_reg/1f1aca01b83efe0a634b85e806fc20a58b10a17ad992bdb62a29b7d07ee574f7.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "1f7e2dbe43b28c69ba29d5113ab0b8554cf9a21fe6de11c6f81f8dcc78045615", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "df19f7a9d185a905d94fd90b73ae5a0510f0ecec42f033ba7dd487a23105d5c0", "generation_seconds_batch": 18.895814412040636, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9054327011108398, "target_z": 0.14941408494653838, "ce": 5.359400749206543, "genuineness": 3.9048471450805664, "identity": 0.6403186321258545, "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_CART_reg/1f7e2dbe43b28c69ba29d5113ab0b8554cf9a21fe6de11c6f81f8dcc78045615.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": "218b58036378e426eaa34f948b45778a0d14191b5a0c35c5bf39b464496985c2", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0503dbbb9a7b44735cafd52058efe999f86c56d3f05568eb0698680613453314", "generation_seconds_batch": 15.96952764200978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2382184267044067, "target_z": -0.3378906535683216, "ce": 6.197352409362793, "genuineness": 4.394378662109375, "identity": 0.4804537296295166, "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_CART_reg/218b58036378e426eaa34f948b45778a0d14191b5a0c35c5bf39b464496985c2.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "219a1090c6c49b2456b755c3dfae075fa87ade5267aab17652b3cb7b72afaf48", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "90073bc4e7cc91d7271b65d6ae5f7a88f7b64730a6098f692df8858b30890098", "generation_seconds_batch": 20.44804618495982, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6045845746994019, "target_z": -0.07031250544158506, "ce": 6.268836498260498, "genuineness": 4.432034492492676, "identity": 0.46292537450790405, "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_CART_reg/219a1090c6c49b2456b755c3dfae075fa87ade5267aab17652b3cb7b72afaf48.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "2245eee84a7069ea580f79bea793a8dc45060460fcfe8ab37211c034e0b74f0b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b9307dd2ccef8d77fe97cf07a9c7543d0b5101fe39338697389139cefe6255e3", "generation_seconds_batch": 21.702343598008156, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0540037155151367, "target_z": 0.9882811765386017, "ce": 5.49045467376709, "genuineness": 3.6349775791168213, "identity": 0.2385353147983551, "hypothesis": "De duur staat nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "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_CART_reg/2245eee84a7069ea580f79bea793a8dc45060460fcfe8ab37211c034e0b74f0b.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "228f4cdbd43c30eef3ca7065c148fb8924511414b123e76e87a1d486825fe2b5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "67e6c63968f32eb3edb1dcd63bd06a550ed31269415c32bcc4b0eaace9767010", "generation_seconds_batch": 21.343929087975994, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.58319091796875, "target_z": -0.08593751632475519, "ce": 5.990842819213867, "genuineness": 4.356721878051758, "identity": 0.52189701795578, "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_CART_reg/228f4cdbd43c30eef3ca7065c148fb8924511414b123e76e87a1d486825fe2b5.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "235d87ca56c38df3f609e5ef74196d76231d2933e68058fd26a9c4c552d1e70b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "775249136c5525ccace20d213aaca9fe409412e81a8ae6ed8efc178d9a96e0e6", "generation_seconds_batch": 20.128580239019357, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8240222930908203, "target_z": 0.8203125925069461, "ce": 5.879645347595215, "genuineness": 4.168440818786621, "identity": 0.49319374561309814, "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_CART_reg/235d87ca56c38df3f609e5ef74196d76231d2933e68058fd26a9c4c552d1e70b.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": "2380340b828f75b3cc72cd56b826b104fee9ca43f024a60ac20c5a15e51fb6fa", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "258326ed15f7b2c7ce01f2c470cf045df659111428e1836b17737b0cf96d36dd", "generation_seconds_batch": 20.227202686015517, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2676347494125366, "target_z": -0.3164062309544523, "ce": 6.1020402908325195, "genuineness": 4.319066047668457, "identity": 0.7266260385513306, "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_CART_reg/2380340b828f75b3cc72cd56b826b104fee9ca43f024a60ac20c5a15e51fb6fa.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "23965bf7e90ccf4f58a3abf4a211d63bfe7047b45b760e268d8677977351156a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5def7b00eb99ee7d38cff597c57c886f17ef69563e98c90c813b3c43a5601444", "generation_seconds_batch": 19.581233522039838, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2437195777893066, "target_z": 0.3964843355485083, "ce": 5.879645347595215, "genuineness": 3.948779582977295, "identity": 0.5961971282958984, "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_CART_reg/23965bf7e90ccf4f58a3abf4a211d63bfe7047b45b760e268d8677977351156a.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "24c127c98da77aa50b94d16f7fd16dc33957b4962d278cf25da832a02191c939", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "99545d9b796ade91a1c4605a315ab6b19371d2964425a27975bf21c1c8d7b28a", "generation_seconds_batch": 21.17533648398239, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3560361862182617, "target_z": 0.4785155773861307, "ce": 5.669164657592773, "genuineness": 4.030367851257324, "identity": 0.015295803546905518, "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_CART_reg/24c127c98da77aa50b94d16f7fd16dc33957b4962d278cf25da832a02191c939.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "25284a49ddfa407f63092119a3e567ed74bad8c8b0bac112cb10a42e9e371625", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a1bb9cf304a872d1d24326aa8cb4b46c1639e2f5d5ef7b45f474578d9bc40def", "generation_seconds_batch": 19.841903843916953, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4763755798339844, "target_z": 0.5664063180198133, "ce": 5.9431867599487305, "genuineness": 3.9048471450805664, "identity": 0.7889411449432373, "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_CART_reg/25284a49ddfa407f63092119a3e567ed74bad8c8b0bac112cb10a42e9e371625.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "2528eeb81534a281dcad06a358d46ae381c3ab496d8a850893e3dcbd3f5599c3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8aa4f9d9576fb1b855f3982fff6123fb2f6fc9fe7b7d3e50407e67ae7d33eb61", "generation_seconds_batch": 15.884939100942574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4282398223876953, "target_z": 0.5312500217663403, "ce": 6.157639026641846, "genuineness": 4.15588903427124, "identity": 0.7094002962112427, "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_CART_reg/2528eeb81534a281dcad06a358d46ae381c3ab496d8a850893e3dcbd3f5599c3.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "25befae6f4e99639b89206e1b3186d6225ef53fa58785f110217e1a31136b61b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "023e70b9e78c19d4cd5afcb9dfda0d205a9b823694b8b661b5d3fb269da61d92", "generation_seconds_batch": 18.93665307399351, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3852996826171875, "target_z": -0.2304688016950581, "ce": 6.316492557525635, "genuineness": 4.432034492492676, "identity": 0.4100740849971771, "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_CART_reg/25befae6f4e99639b89206e1b3186d6225ef53fa58785f110217e1a31136b61b.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "2607928510259d54a45aa4c60f3456364151ddc27b61fa0f55f5a69d6013da11", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "cb9fd98f6f8eb7086e556d338d47b2a936f3287e493fa89b94c28f2aee47a8ce", "generation_seconds_batch": 19.7845632720273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.791931629180908, "target_z": 0.7968749455841494, "ce": 5.319687366485596, "genuineness": 3.6663577556610107, "identity": 0.048263370990753174, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/2607928510259d54a45aa4c60f3456364151ddc27b61fa0f55f5a69d6013da11.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": "26139a1f47f14877af27137a14c24874a920df973a0f8cd84e52ec8374690f00", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3581f83014fdd7ecac62dd2d585916581775e13fd256a7c9671b459acb37aa7c", "generation_seconds_batch": 15.076472377055325, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.631173849105835, "target_z": -0.7812500217663403, "ce": 6.0305562019348145, "genuineness": 4.055472373962402, "identity": 0.399746835231781, "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_CART_reg/26139a1f47f14877af27137a14c24874a920df973a0f8cd84e52ec8374690f00.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "264d8273572f0a6890030eeabfd915fd22fcd2f5d6ec10b40e61cd26ac82a4f6", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "271554dad38fdf3fcd6067a0bcebf93e0de2e8525b65a5e2c31e130f7bb6cbaa", "generation_seconds_batch": 19.489968045963906, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.230348587036133, "target_z": 0.3867187363960373, "ce": 6.133810997009277, "genuineness": 4.042920112609863, "identity": 0.2692357003688812, "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_CART_reg/264d8273572f0a6890030eeabfd915fd22fcd2f5d6ec10b40e61cd26ac82a4f6.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "268a1f4fcf04cf9817220f8ecff991ca791f73deac1f72c20151ba557ddd45c0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "772125ae3c5bfb3c12fbff40aaed84134eeeb40d5e748d272348d3974838a0f9", "generation_seconds_batch": 19.36752060998697, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.536392331123352, "target_z": -0.12011720042376453, "ce": 6.260893821716309, "genuineness": 4.049196243286133, "identity": 0.10531394183635712, "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_CART_reg/268a1f4fcf04cf9817220f8ecff991ca791f73deac1f72c20151ba557ddd45c0.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "a625f93a23b194ed4c9fad28bf2267d66b223c2992286e2757ce342e5564e26a", "generation_seconds_batch": 15.879502214025706, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6900063753128052, "target_z": -0.7382812636039626, "ce": 6.395919322967529, "genuineness": 3.998987913131714, "identity": 0.1464369297027588, "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_CART_reg/26d12995a5b024b42cbe037fdee48ffb67b79df656c09f059592185aa881aca9.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "26ece3889e8d9fb23821164eb4520949b4ec9a790b124d9f71c05c5eeca2c108", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "4ab2a20ade27370862f1abe3298e4732022527a18121cf4a23a29e787c7641d9", "generation_seconds_batch": 20.191695200977847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.814509630203247, "target_z": 0.0830077669267251, "ce": 6.078212261199951, "genuineness": 4.425758361816406, "identity": 0.6183395385742188, "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_CART_reg/26ece3889e8d9fb23821164eb4520949b4ec9a790b124d9f71c05c5eeca2c108.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "270097de4f9ebfce7fa24fd0d9261e07850d4a48bb7c21f8f965f6d6321f760a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1877de1ecbf2395d1ed89b1a9a65134c774eefdd825c00775f016ad636cd9e65", "generation_seconds_batch": 20.275696104974486, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9843217134475708, "target_z": 0.2070312418376224, "ce": 5.9431867599487305, "genuineness": 4.13078498840332, "identity": 0.5825536251068115, "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_CART_reg/270097de4f9ebfce7fa24fd0d9261e07850d4a48bb7c21f8f965f6d6321f760a.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "2706f9b8d479b449e92197285ab75ecbcfc8b170bd39530549de7f3578c126d8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1bd8861a82bf7212aa82ee8d620b13d9210281f537824338432f55a7f92bc63b", "generation_seconds_batch": 22.161305128014646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.059046745300293, "target_z": -0.46874997823365977, "ce": 6.332377910614014, "genuineness": 4.444586753845215, "identity": 0.355999231338501, "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_CART_reg/2706f9b8d479b449e92197285ab75ecbcfc8b170bd39530549de7f3578c126d8.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "273cabdc5444520a46044d54260a5c61f463448c3827104e19e5ccf8e9d254b6", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e3b59a9106102b395b7ebc1e35abcc0744ff29296d4aab9f6def756fc90d9fa5", "generation_seconds_batch": 17.269429969019257, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5138144493103027, "target_z": 0.5937500652990207, "ce": 5.645336627960205, "genuineness": 4.030367851257324, "identity": 0.011125186458230019, "hypothesis": "Die deur 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_CART_reg/273cabdc5444520a46044d54260a5c61f463448c3827104e19e5ccf8e9d254b6.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "2741a429eda2c92bfba0c29d4fbb289083441e02adeb3b2e22a33298e6b6a034", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "08002ecf12b98a5c6e96bbe33934a3859819a2ad9abe4cf221093b81d536b246", "generation_seconds_batch": 20.191695200977847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9910072088241577, "target_z": 0.2119140414138579, "ce": 5.95112943649292, "genuineness": 4.149612903594971, "identity": 0.47829797863960266, "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_CART_reg/2741a429eda2c92bfba0c29d4fbb289083441e02adeb3b2e22a33298e6b6a034.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "279d98c1958c490badf18ccd8ed7bb46019c386122f98043453e50c969e120dd", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e8281b9f4fbdea6337d11125e2c034ba49030f8a962c184d311e2a55c30ce79f", "generation_seconds_batch": 20.386883034021594, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3212716579437256, "target_z": 0.45312505441585066, "ce": 5.974957466125488, "genuineness": 4.256305694580078, "identity": 0.4152711033821106, "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_CART_reg/279d98c1958c490badf18ccd8ed7bb46019c386122f98043453e50c969e120dd.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "27df7eec2285930bb77bd10bbf66f9adc7f34c1ecb8fc567c66e0f47051d9644", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e2dab3b90f1f2d4410ed78a22e706856afd40fdac6617109e975286a6a111bb2", "generation_seconds_batch": 23.924444830045104, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4335882663726807, "target_z": 0.535156296253473, "ce": 6.1894097328186035, "genuineness": 4.055472373962402, "identity": 0.6011767387390137, "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_CART_reg/27df7eec2285930bb77bd10bbf66f9adc7f34c1ecb8fc567c66e0f47051d9644.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "28b5674d98a544bbfb899c7d58da96a49e45b317c1b1c907f6a048ac8ac4a4a9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "10ac39000c688cf6bea9e58b72c9381391a2373ce74e488c5555463ab264d0e0", "generation_seconds_batch": 21.00786956003867, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3854525089263916, "target_z": 0.5, "ce": 5.919358730316162, "genuineness": 4.18099308013916, "identity": 0.500691294670105, "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_CART_reg/28b5674d98a544bbfb899c7d58da96a49e45b317c1b1c907f6a048ac8ac4a4a9.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "28ebc8fe7576cf039e85eac49e8cd0629a0081237442f1aaf40755398e4b0c01", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "213e507a9f3d18cb92424da84be7897f2d3668c19aef1921749749e9c0fc4332", "generation_seconds_batch": 24.2782056289725, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0058679580688477, "target_z": 0.9531248802851287, "ce": 5.772419452667236, "genuineness": 3.860914945602417, "identity": 0.3484576344490051, "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_CART_reg/28ebc8fe7576cf039e85eac49e8cd0629a0081237442f1aaf40755398e4b0c01.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "28f6e8bd3ace2178dde2868c22fb52023117bb6aff92ded9380996318ade60fe", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0c14302d107a91d35cf03c74e2e26e2d5d9f1b4650b006aebbd348bae89480d4", "generation_seconds_batch": 21.88251304801088, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8867132663726807, "target_z": 0.13574221130693465, "ce": 6.086154937744141, "genuineness": 4.068024158477783, "identity": 0.37623172998428345, "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_CART_reg/28f6e8bd3ace2178dde2868c22fb52023117bb6aff92ded9380996318ade60fe.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "293cd02915867b00785914aa262ea19baa51477d84ef4119132bda80326c2594", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b588224659630b4f53618dd24a8cf063b1da23eb53cadf79d97dad1f211bced0", "generation_seconds_batch": 21.463669580989517, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3719286918640137, "target_z": -0.24023440084752906, "ce": 5.8875885009765625, "genuineness": 4.287685871124268, "identity": 0.5516273975372314, "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_CART_reg/293cd02915867b00785914aa262ea19baa51477d84ef4119132bda80326c2594.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": "293df17eba28ccbbe0343ca5217b1bc2f82cf4269264d9daba41c17e1e210063", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ef14d192a22339646b9d9164f08416b22bb84253f8cf113e320fcd51b75ed575", "generation_seconds_batch": 14.846603980055079, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4628517627716064, "target_z": -0.17382808282771575, "ce": 5.998785495758057, "genuineness": 3.7040140628814697, "identity": 0.5162261128425598, "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_CART_reg/293df17eba28ccbbe0343ca5217b1bc2f82cf4269264d9daba41c17e1e210063.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "29d4f3d7fe6185cae9aa698d428a6c1d13c9bc06df9736b49ae7a7b4bc02c5a8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d2fa561fb368ca21cbeec43551cafaf3f2c2e0f4ee97beb562546a87cc633e08", "generation_seconds_batch": 15.010764572070912, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7583513259887695, "target_z": 0.04199214600791389, "ce": 5.990842819213867, "genuineness": 4.243753433227539, "identity": 0.5181930661201477, "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_CART_reg/29d4f3d7fe6185cae9aa698d428a6c1d13c9bc06df9736b49ae7a7b4bc02c5a8.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "2aa87a419e263a9111b0515f85f9b983744e2d37cfd5977aa7d3d08c9edea488", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "13e94541614e502cfa2e7d2feeb839dcd64ff24aab6081544202ea2520571b19", "generation_seconds_batch": 21.476005648029968, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6741138696670532, "target_z": -0.019531285370302912, "ce": 6.054384231567383, "genuineness": 4.15588903427124, "identity": 0.16070963442325592, "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_CART_reg/2aa87a419e263a9111b0515f85f9b983744e2d37cfd5977aa7d3d08c9edea488.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "2b22f303bfecbd0491226b5ecef289a38587e6b95af18dde4b59d2ef21f5b3b9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "62bbfa409be2e606f5d75a04f5436ba80a2c3031647b0c3c3446823078f325f1", "generation_seconds_batch": 21.17533648398239, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4656786918640137, "target_z": 0.5585937690455477, "ce": 5.178704738616943, "genuineness": 3.735394239425659, "identity": 0.42577308416366577, "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_CART_reg/2b22f303bfecbd0491226b5ecef289a38587e6b95af18dde4b59d2ef21f5b3b9.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "2b3e179b3fed74375225c5988e227045216afda9b2383b1614d0f2fa027cd5e7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "41f6213770d4ac004993ea88da4df690b7c6a0a8a8aa6b6743f35df35381ea4c", "generation_seconds_batch": 20.60425872390624, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8399146795272827, "target_z": 0.10156252720792532, "ce": 6.268836498260498, "genuineness": 4.482243061065674, "identity": 0.74480140209198, "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_CART_reg/2b3e179b3fed74375225c5988e227045216afda9b2383b1614d0f2fa027cd5e7.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "2bd51fd07810262e95541a565e4df8cc47945e93206b7987dd3f15780fea8920", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9eafd65100ed5557133d3c1b60f4a159ffe0df4081e62f668ddc786be6d63058", "generation_seconds_batch": 19.08817491203081, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3264672756195068, "target_z": -0.2734374727920747, "ce": 6.38003396987915, "genuineness": 4.444586753845215, "identity": 0.3704068660736084, "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_CART_reg/2bd51fd07810262e95541a565e4df8cc47945e93206b7987dd3f15780fea8920.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "2c0a2900c0f20e8f3f314133bdc74d498268d45365e0d0f6081813c7c7426ff3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "322e5d26d6b558acf7b27467b2b9190a170896a5109b2c091bdf6496ec3cecb8", "generation_seconds_batch": 14.795922335004434, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9627753496170044, "target_z": -0.5390624836752448, "ce": 6.24500846862793, "genuineness": 4.507347106933594, "identity": 0.46212947368621826, "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_CART_reg/2c0a2900c0f20e8f3f314133bdc74d498268d45365e0d0f6081813c7c7426ff3.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "2c57a09bc75865720c0debd13ba9cb36678d129854c736fb8db453165ab05bda", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f0c5ca21c36f135558ccfdea68359715e2f03ded4c1bacc94f8511625d272ae1", "generation_seconds_batch": 19.36752060998697, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6687655448913574, "target_z": -0.023437472792074684, "ce": 6.395919322967529, "genuineness": 4.394378662109375, "identity": 0.4803692698478699, "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_CART_reg/2c57a09bc75865720c0debd13ba9cb36678d129854c736fb8db453165ab05bda.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": "2c821dadf430ca0f0b88404c5fedf445951390ff14164f38c1ef27ce77852263", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I wrote the number on the envelope.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0ae44312d191450ee9b672f3cff5221658bd715a13e7c51a751f8561d287f28c", "generation_seconds_batch": 20.975421852082945, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9788205623626709, "target_z": -0.5273437472792075, "ce": 6.3641486167907715, "genuineness": 3.785602569580078, "identity": 0.3334566354751587, "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_CART_reg/2c821dadf430ca0f0b88404c5fedf445951390ff14164f38c1ef27ce77852263.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "2c90dda88dabb6232edafdd6395a397e2e09a9d24d3ce89683391e4cab676fb9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fa713b6d0e03596e4910496ddfbcccfc83bd9fe261b7e18380e40e266e8f5d30", "generation_seconds_batch": 18.970078411046416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8238694667816162, "target_z": 0.08984379081188798, "ce": 6.23706579208374, "genuineness": 4.099404335021973, "identity": 0.5297018885612488, "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_CART_reg/2c90dda88dabb6232edafdd6395a397e2e09a9d24d3ce89683391e4cab676fb9.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "2d0baf6a1ae5d28dbc87104f77eefe0b4145478ba1a81fbe3e216d740806eb2a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2e48412b553415836bcb8ed245a45a055312feb3175fc767756ae83223ea3962", "generation_seconds_batch": 20.389921457972378, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1018340587615967, "target_z": -0.4374999564673195, "ce": 6.229123115539551, "genuineness": 4.306513786315918, "identity": 0.6969471573829651, "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_CART_reg/2d0baf6a1ae5d28dbc87104f77eefe0b4145478ba1a81fbe3e216d740806eb2a.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "2d7d68fea0c6d32e77f21b9b885478b75ca729bf683c5eaf338e11a8597c34b8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "1571accd6d7d20cd079a4ffa8d2300740862d309723a8dc09f003b2a1ee580f2", "generation_seconds_batch": 21.38856685499195, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.663569927215576, "target_z": 0.7031250544158506, "ce": 5.299830436706543, "genuineness": 3.8671910762786865, "identity": 0.20693561434745789, "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_CART_reg/2d7d68fea0c6d32e77f21b9b885478b75ca729bf683c5eaf338e11a8597c34b8.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "2e19424c105f60277961c0bbb2cd856df230cd05ae2c8ca6999b88721c1f0907", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "8dfd434a8653ec4be637b89d50b97ac21ab3019a6bfdd8d0384414740ce9ca0f", "generation_seconds_batch": 14.821375350933522, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1767117977142334, "target_z": -0.38281246190890456, "ce": 5.808161735534668, "genuineness": 4.256305694580078, "identity": 0.5946750640869141, "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_CART_reg/2e19424c105f60277961c0bbb2cd856df230cd05ae2c8ca6999b88721c1f0907.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "2e25123c40d707aa7caae88e491f22e89168f7759ab1ac2897d8b0ecf8f7035e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9da2c1f8e1672f8a72acd8d17e47bcd55828e8b5228d7445b7348c527c7af138", "generation_seconds_batch": 21.806784193962812, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3373167514801025, "target_z": 0.46484370374652695, "ce": 5.998785495758057, "genuineness": 3.672633647918701, "identity": 0.418327271938324, "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_CART_reg/2e25123c40d707aa7caae88e491f22e89168f7759ab1ac2897d8b0ecf8f7035e.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "2e74e714568c985098ad0699cd9ceca9740cb2d022a9583be0272d7e316e4cd7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d0300ddf93b4975348d3a09d71a0bad077bbf3d57d99a9bc593b6fe6ade2395d", "generation_seconds_batch": 20.161289564915933, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.59938907623291, "target_z": 0.6562501088317013, "ce": 5.8875885009765625, "genuineness": 4.042920112609863, "identity": 0.5966912508010864, "hypothesis": "La porte est encore ouverte ! J'ai posé ton manteau sur la chaise ! Dis-moi, dis-maître potes ! Dis-moi quand tu es prêt ?", "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_CART_reg/2e74e714568c985098ad0699cd9ceca9740cb2d022a9583be0272d7e316e4cd7.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "2e92317914d1a6e222d121f5f8e04e11f5a5a45b6438da75e1063755c31dccdb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8bc606f7bc89eaa612b95c4e21c88753aac402d914e171748798e54956551aeb", "generation_seconds_batch": 19.72106581600383, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.567298412322998, "target_z": 0.6328124619089046, "ce": 5.534139156341553, "genuineness": 3.9801597595214844, "identity": 0.294064462184906, "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_CART_reg/2e92317914d1a6e222d121f5f8e04e11f5a5a45b6438da75e1063755c31dccdb.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": "2eba41e279a764969476935451513466c8ff93141cbb9b6a2dd2e4332db1e59e", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3e0a3b44f02c85bd738452f6edcf607b330d79b9ab7418a8df0bc9ba649c2dcd", "generation_seconds_batch": 21.10331853700336, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9520785212516785, "target_z": -0.5468749891168299, "ce": 6.197352409362793, "genuineness": 4.256305694580078, "identity": 0.6223265528678894, "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_CART_reg/2eba41e279a764969476935451513466c8ff93141cbb9b6a2dd2e4332db1e59e.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "2ef44b6cbfe24040821b340772dfdb0680bf758d1a1fec362e763cd1b8e0044d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4972e0f78b0e8a10687b0774ae9d6fa29579c38dff5ea53b4af277899f5a6ed7", "generation_seconds_batch": 19.008948259055614, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.684810757637024, "target_z": -0.011718736396037342, "ce": 6.332377910614014, "genuineness": 4.400654315948486, "identity": 0.4482346177101135, "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_CART_reg/2ef44b6cbfe24040821b340772dfdb0680bf758d1a1fec362e763cd1b8e0044d.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "2eff4d6008167ad865a1d1cf957b47e22bac4ccfa64feb91eb450272301fdd8a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ac8c2c601008e0be265022abba5636fc3f206e4275979302575f94cc5368f62f", "generation_seconds_batch": 15.20800564205274, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7463175058364868, "target_z": 0.03320315900990665, "ce": 6.268836498260498, "genuineness": 4.375550270080566, "identity": 0.43657225370407104, "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_CART_reg/2eff4d6008167ad865a1d1cf957b47e22bac4ccfa64feb91eb450272301fdd8a.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "2f3d114e29e8b4e596797c3f9f338a0db12d18820365fe998b2d9e9cf74a0979", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "307cbe8c599954ea223058955bef35ffc96915d908bd7a6979e2938387261cb5", "generation_seconds_batch": 20.78243125299923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5230213403701782, "target_z": -0.12988279957623547, "ce": 6.3641486167907715, "genuineness": 4.331617832183838, "identity": 0.4810960292816162, "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_CART_reg/2f3d114e29e8b4e596797c3f9f338a0db12d18820365fe998b2d9e9cf74a0979.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "2f4480909b582dcf98dc6c8fb0bb01b297dcd71c2c70450c548bc98974d1802c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "484866c91dcaa740a271d2ad174bb64c327608ba432891d90156c8a6f21a336f", "generation_seconds_batch": 22.94136196898762, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6490432024002075, "target_z": -0.03784182731386645, "ce": 6.260893821716309, "genuineness": 4.457139015197754, "identity": 0.449615478515625, "hypothesis": "J'ai trouvé la la petite boîte. Elle était près de 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_CART_reg/2f4480909b582dcf98dc6c8fb0bb01b297dcd71c2c70450c548bc98974d1802c.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": "30192b351eb73ae27a75139fff75bfc87f4c247013b5ea87355ca90870fb9a4c", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eddb1e028c1e6fc03399367afe9f199b1c113d3e6d46922e733c27795529e578", "generation_seconds_batch": 15.426801213994622, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1820602416992188, "target_z": -0.3789061874217718, "ce": 6.205295085906982, "genuineness": 4.256305694580078, "identity": 0.7005804777145386, "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_CART_reg/30192b351eb73ae27a75139fff75bfc87f4c247013b5ea87355ca90870fb9a4c.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "30a2b3d8701bd452e6da5e899c5981a22705dbeb7a7a85bf48b016b0cc9b02c9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4c5a8ae7c08fb9e24aaf700412120bc024c25ed34b7a0fe7c3a498568380ec47", "generation_seconds_batch": 15.050687537062913, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6888220310211182, "target_z": -0.00878907406336826, "ce": 6.268836498260498, "genuineness": 4.149612903594971, "identity": 0.4597257673740387, "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_CART_reg/30a2b3d8701bd452e6da5e899c5981a22705dbeb7a7a85bf48b016b0cc9b02c9.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "31976e27c6d3560a45fa01cf35b84d236fd8d7c5e98ed9684402fbf5df62835a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f5e145fe320956ee752f6de6a0e9d2ff0ecbf6b840ba682d31b8095a4ec80e92", "generation_seconds_batch": 20.265794175909832, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5069761276245117, "target_z": -0.14160153597227282, "ce": 6.109982967376709, "genuineness": 4.595211505889893, "identity": 0.67900550365448, "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_CART_reg/31976e27c6d3560a45fa01cf35b84d236fd8d7c5e98ed9684402fbf5df62835a.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "31a268cb7621aa2b9f8106f1217b2eb1c272fe30b1a37252546a16882973b40c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7f777709268a2fcb7728607be1b1b0ae91413eeff00a0b622b3f87ee4207aca7", "generation_seconds_batch": 20.60425872390624, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6500461101531982, "target_z": -0.03710934643167842, "ce": 6.221180438995361, "genuineness": 4.570107460021973, "identity": 0.5593018531799316, "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_CART_reg/31a268cb7621aa2b9f8106f1217b2eb1c272fe30b1a37252546a16882973b40c.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "31df3fe10adbf7c965f2e5b2e8386d09bb8327584a9d643e822c4078aac11217", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0dd9907dee02f2c591018e31f4f3338b62a8dc1649b5c2703b9c4209cfdd3a84", "generation_seconds_batch": 23.003980839974247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.040480136871338, "target_z": 0.24804694982179462, "ce": 6.181467056274414, "genuineness": 4.306513786315918, "identity": 0.42772966623306274, "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_CART_reg/31df3fe10adbf7c965f2e5b2e8386d09bb8327584a9d643e822c4078aac11217.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "3231a544d7a26f5138558c10530f77f6249fabba00bf703c45b2c8ded675d656", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3215bc24cda4f2cdc5670b22bedc811dc533b47abc301ca4700dacebd87dfb68", "generation_seconds_batch": 22.62403995904606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.706357002258301, "target_z": 0.7343749020514688, "ce": 5.919358730316162, "genuineness": 3.7416701316833496, "identity": 0.574532151222229, "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_CART_reg/3231a544d7a26f5138558c10530f77f6249fabba00bf703c45b2c8ded675d656.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "3290090f8adba8b104c5d0d948848e0a5ee3278615697fdfd3b376b74ab92a58", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1e50c4d44ccb8c2a6b8142b7088a6cf9b9f9cd00819aa14a2ea05e665f5bdaf1", "generation_seconds_batch": 21.76622396602761, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9762990474700928, "target_z": 0.20117183010692322, "ce": 5.879645347595215, "genuineness": 3.8420867919921875, "identity": 0.34892207384109497, "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_CART_reg/3290090f8adba8b104c5d0d948848e0a5ee3278615697fdfd3b376b74ab92a58.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "32ba1e8cd3bd02cbfebefaf8bb66703ae24911e29f40301853e79efe05179195", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "42bb87a0733027c802520d654faaf27cb2881945952ea5243ccb96e1afce97a1", "generation_seconds_batch": 25.389631084050052, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4335882663726807, "target_z": 0.535156296253473, "ce": 5.9431867599487305, "genuineness": 4.331617832183838, "identity": 0.556342363357544, "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_CART_reg/32ba1e8cd3bd02cbfebefaf8bb66703ae24911e29f40301853e79efe05179195.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "330ae9fb7caf250e7aa50ec803c2f3a4eba3e08436bf560afdf39599f2fd42a5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ffaf44f3c62ef6ce30cda9099e5435e01c5d09a098d6d5547f045e469d9bb326", "generation_seconds_batch": 18.985369650064968, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5069761276245117, "target_z": -0.14160153597227282, "ce": 6.284721851348877, "genuineness": 4.20609712600708, "identity": 0.2944833040237427, "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_CART_reg/330ae9fb7caf250e7aa50ec803c2f3a4eba3e08436bf560afdf39599f2fd42a5.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "331137ace4a926cf90dc747b934cdad9732b800308377b6b76401393a7b95275", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ddce9433d28240c5119b1211fa368008887ad5a0da1a3e190c79457990d908a8", "generation_seconds_batch": 19.347513125045225, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3480136394500732, "target_z": 0.47265625272079254, "ce": 5.871703147888184, "genuineness": 3.8797430992126465, "identity": 0.369659960269928, "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_CART_reg/331137ace4a926cf90dc747b934cdad9732b800308377b6b76401393a7b95275.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "3316766d64633e64fd086e9bef7f7718b5a6c2d81a2e5af00db3c8eac9e2a27c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fc00c313054881b8967870006ed6a964ce7e34af6a1fa8628f6bc1e33480d99b", "generation_seconds_batch": 21.028623520978726, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7275980710983276, "target_z": 0.019531285370302912, "ce": 6.149696350097656, "genuineness": 4.243753433227539, "identity": 0.4147809147834778, "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_CART_reg/3316766d64633e64fd086e9bef7f7718b5a6c2d81a2e5af00db3c8eac9e2a27c.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "332595d4591c95bc5d6c0a4901a49e933c35d7b617b0d35c3bc7cd9d4a16ea03", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "efdd9d657721a09f4814425b922d4dfb0778113fb722dbf66ec672f9a6db26ea", "generation_seconds_batch": 19.8977074889699, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3959965705871582, "target_z": -0.22265625272079254, "ce": 6.213237762451172, "genuineness": 4.381826400756836, "identity": 0.367858350276947, "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_CART_reg/332595d4591c95bc5d6c0a4901a49e933c35d7b617b0d35c3bc7cd9d4a16ea03.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "3354703750c26af04790fd19fa55c422c859a2ad0d065116807e5fcd0af74e27", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ad1334bf96b64c4a40eed49757c3897b3319ccec5c8f7cf439e61fddc0996ca5", "generation_seconds_batch": 20.91190315701533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4814183712005615, "target_z": -0.8906250108831701, "ce": 6.340320587158203, "genuineness": 4.519899368286133, "identity": 0.6596493721008301, "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_CART_reg/3354703750c26af04790fd19fa55c422c859a2ad0d065116807e5fcd0af74e27.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "33963bcc4a7b02696d8de62ba09bf3e9d5710f62ce81d1cff2edb931da40db40", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ef7be30a559cf26613e3bd678708c7bf9ebb5492a3f2827c1ac7f7fe94b2bd5f", "generation_seconds_batch": 20.10162755602505, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.829217791557312, "target_z": 0.09374997823365974, "ce": 5.847874641418457, "genuineness": 4.18726921081543, "identity": 0.64329993724823, "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_CART_reg/33963bcc4a7b02696d8de62ba09bf3e9d5710f62ce81d1cff2edb931da40db40.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "33b090222f34f1b304f6b75c7b9498bf86bc96429eb1edd7721f7f42b6c36d55", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "93f31b31c7acfe40899494c219123ddb3890694bd769dc3646d100fd0d5b8c65", "generation_seconds_batch": 21.510647344985045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9948657751083374, "target_z": -0.5156250108831701, "ce": 6.022613525390625, "genuineness": 4.557555198669434, "identity": 0.6346968412399292, "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_CART_reg/33b090222f34f1b304f6b75c7b9498bf86bc96429eb1edd7721f7f42b6c36d55.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": "33b718959db70d73ccaaeb0e7db624e14e57d30403125dd70fb5edb79a091fc8", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I wrote the number on the envelope.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a603067d58a9f590c569ee51417fc907a1651b8762819ca60e8eb2a24a630bef", "generation_seconds_batch": 15.302311728941277, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8290650844573975, "target_z": -0.6367187363960374, "ce": 6.24500846862793, "genuineness": 3.8922953605651855, "identity": 0.5033048391342163, "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_CART_reg/33b718959db70d73ccaaeb0e7db624e14e57d30403125dd70fb5edb79a091fc8.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "6737c5357bbf47c55252b974d7496ab5e7e0a78f37a3fd95ad56a9a3dc922c4d", "generation_seconds_batch": 14.931478124926798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.2781788110733032, "target_z": -1.039062483675245, "ce": 6.284721851348877, "genuineness": 4.469690799713135, "identity": 0.4554497003555298, "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_CART_reg/33f63538ac2c49676cc53462aa2024b47e58cb29b8690774a6580785ecb0e60e.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "34459d8455e11d83bae2b2389743660007faf6ed65c44a5b89a1bfee18ab0154", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2a4c89518b0c6568b0d69aaeabc012887e4827e870c81e9edb17dbdc876f45d7", "generation_seconds_batch": 19.53446306101978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6620800495147705, "target_z": -0.028320272368310157, "ce": 6.324435234069824, "genuineness": 4.306513786315918, "identity": 0.6213858723640442, "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_CART_reg/34459d8455e11d83bae2b2389743660007faf6ed65c44a5b89a1bfee18ab0154.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "3454cd3bd0f59de2d0a2241d5edf380ef41126748ee689768f920377ed65f9b9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a7517458091eb6f7bfd58fb841b6e56474b72c074fc50fbc5dfac0e2165003e5", "generation_seconds_batch": 19.832552331034094, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8076714277267456, "target_z": -0.6523437472792075, "ce": 6.292664527893066, "genuineness": 4.068024158477783, "identity": 0.4488878846168518, "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_CART_reg/3454cd3bd0f59de2d0a2241d5edf380ef41126748ee689768f920377ed65f9b9.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "ccc97b3614e3ead1e246e8f1fb3304e90bd6244a64586436b2b1303f07a70248", "generation_seconds_batch": 16.305065630935133, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6739611625671387, "target_z": -0.75, "ce": 6.443574905395508, "genuineness": 4.13078498840332, "identity": 0.19054603576660156, "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_CART_reg/347b084eb1c87968a3fc14b855043a2de140e81cf0874b7d65525285bfb31414.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "3502654cca80df02b0d8144653738b73081639a067d98082e6757443ab0602fc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "adca3c8df635e1e0fca4267f5edecc5016462b07ce78d60fafeca729121a89b8", "generation_seconds_batch": 21.191109837032855, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.166167736053467, "target_z": 0.339843790811888, "ce": 5.935244083404541, "genuineness": 3.911123275756836, "identity": 0.5533865094184875, "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_CART_reg/3502654cca80df02b0d8144653738b73081639a067d98082e6757443ab0602fc.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "3534bf260e1f42a099af40cd5ceab4bcc7078831a32c67c9a6b83916de3bb84c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "bc96ae84e7e250c554025b67083bc29eabb9614f09fa070139870b04f520e0ad", "generation_seconds_batch": 15.73440755798947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.834413468837738, "target_z": -0.6328125054415851, "ce": 6.1020402908325195, "genuineness": 4.381826400756836, "identity": 0.6136513352394104, "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_CART_reg/3534bf260e1f42a099af40cd5ceab4bcc7078831a32c67c9a6b83916de3bb84c.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "3581aa27ba643a1a49cd0a80010caaf52e24cf958c67ca24498fe728763f229f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d13a3e4421ddd3393cf45c5e363cc103203e03db80325c846da320af591bc4a0", "generation_seconds_batch": 21.848669525934383, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4175429344177246, "target_z": 0.5234374727920746, "ce": 6.038498878479004, "genuineness": 4.017816066741943, "identity": 0.5840924978256226, "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_CART_reg/3581aa27ba643a1a49cd0a80010caaf52e24cf958c67ca24498fe728763f229f.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "35d5437be45c1eba48926466538719c775dfaff51e5d0a46cac263a27389618f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "19be9c3b2a758316169f328842ece061d19cdf403166e9530aea734204e336d5", "generation_seconds_batch": 21.274790597963147, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9841689467430115, "target_z": -0.5234375163247552, "ce": 6.117925643920898, "genuineness": 4.20609712600708, "identity": 0.47063714265823364, "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_CART_reg/35d5437be45c1eba48926466538719c775dfaff51e5d0a46cac263a27389618f.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "35e4bd8947930c04bfaee444e2714c7dbb26736d20c060a3c2aab4509244b01d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b849ba60e18124f057199f235712b3cecc319fa2fd259be7eb0b19a47392ebd3", "generation_seconds_batch": 20.023067054105923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.551253318786621, "target_z": 0.6210938125782283, "ce": 5.998785495758057, "genuineness": 4.193545341491699, "identity": 0.5051219463348389, "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_CART_reg/35e4bd8947930c04bfaee444e2714c7dbb26736d20c060a3c2aab4509244b01d.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "360826cd8e6e6b0c5f8d543851be49d157f7cd2705f67f29e5250004d2917fc5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ba578902355115f326929ddbb4e769944d01bc9fc1b30f6a6efd9f6e7ad92ca5", "generation_seconds_batch": 16.208993748994544, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5751683712005615, "target_z": -0.09179684099009336, "ce": 6.260893821716309, "genuineness": 4.557555198669434, "identity": 0.7413141131401062, "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_CART_reg/360826cd8e6e6b0c5f8d543851be49d157f7cd2705f67f29e5250004d2917fc5.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "372499f6dd169fbece3581d3f3d06380b94f3e5d646aca479dd6868abb102de3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "13c472025190ea2fb7846a60bcf693e97224f4fd46b6d256c92f0c2a3350ae65", "generation_seconds_batch": 18.98300054098945, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3077478408813477, "target_z": -0.2871093464316784, "ce": 6.141753673553467, "genuineness": 4.369274139404297, "identity": 0.35623323917388916, "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_CART_reg/372499f6dd169fbece3581d3f3d06380b94f3e5d646aca479dd6868abb102de3.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "376bcdbe551feb6e8b4e37eb5ffba4d3b0347f8fff60f7726495acb75dfd7b8a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "41c7e17e5ae27c6a3380f87721657d9d40bb0d1cfc4facc4a878465896607a46", "generation_seconds_batch": 19.8963846139377, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.75449275970459, "target_z": 0.7695311983049419, "ce": 5.7128496170043945, "genuineness": 4.017816066741943, "identity": 0.3475305438041687, "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_CART_reg/376bcdbe551feb6e8b4e37eb5ffba4d3b0347f8fff60f7726495acb75dfd7b8a.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "3798d944098b0e753b87b1cc50522717a138e03fd7e6e0c2f8fc25477792b0bb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f363e04fab38feefe355d3ece292adaec6b0f661140df0f900b61b7a1d28f78f", "generation_seconds_batch": 17.269429969019257, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7812347412109375, "target_z": 0.7890623966098838, "ce": 5.498397350311279, "genuineness": 4.017816066741943, "identity": 0.08411125838756561, "hypothesis": "De deur staat nog open! Ik heb je jas op de stoel gelegd. Zeg maar wanneer je wanneer je klaar bent.", "wer": 0.10526315789473684, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/3798d944098b0e753b87b1cc50522717a138e03fd7e6e0c2f8fc25477792b0bb.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "37af09dea4d12c54fbfb3439cb7844ef8a4d468b3db55f4b3a9127b6e8c032d1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4f3734e722aa26776810da8815d4fd97c998fba22a0c7a434d20ffef5e74ef4d", "generation_seconds_batch": 21.779950855998322, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.496279239654541, "target_z": -0.14941408494653838, "ce": 6.149696350097656, "genuineness": 3.6977379322052, "identity": 0.333843469619751, "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_CART_reg/37af09dea4d12c54fbfb3439cb7844ef8a4d468b3db55f4b3a9127b6e8c032d1.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "37f7eec761772acfd6ae73a147ddf220197fdd12b33868d580d8b49fa0f62e0d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4d249f52cbae1a14c87738e9158276c4dff899e1ac699b573e3f54dc1671731f", "generation_seconds_batch": 24.2782056289725, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.519162654876709, "target_z": 0.5976561656554316, "ce": 5.518253803253174, "genuineness": 4.030367851257324, "identity": 0.48789769411087036, "hypothesis": "Te he encontrado la cajita. Estaba junto a la ventana, mi hija. Podemos mirar juntos lo que hay dentro.", "wer": 0.17647058823529413, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/37f7eec761772acfd6ae73a147ddf220197fdd12b33868d580d8b49fa0f62e0d.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "3823eef2085c918224d636edb2a0d0fa8919e5dcf13adeabf984ab0fb5ed8d0b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ab686b659580448e9fb6e0021f4ce17b7e5ea8dc8b6f588245033eb1d86a004a", "generation_seconds_batch": 20.023067054105923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5084660053253174, "target_z": 0.589843790811888, "ce": 5.5539960861206055, "genuineness": 4.017816066741943, "identity": 0.5031347870826721, "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_CART_reg/3823eef2085c918224d636edb2a0d0fa8919e5dcf13adeabf984ab0fb5ed8d0b.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "389089abc9b72d5c82ea3ef712b2c6ba5e156f4cdddb095be72dd9c1f3a57d60", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b556356ca5b1c52541626195616dd669a41d8706156dd4f15a7b3c2289fa0afa", "generation_seconds_batch": 15.195879185106605, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0324573516845703, "target_z": 0.24218745102573444, "ce": 5.9431867599487305, "genuineness": 4.218649387359619, "identity": 0.04522869363427162, "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_CART_reg/389089abc9b72d5c82ea3ef712b2c6ba5e156f4cdddb095be72dd9c1f3a57d60.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "38943f8ae18f8988be06e530ec4799dccf105fa0f9f222c3d418b283bda80542", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "8135e25565548a84db531a124145f4f588eb6c7e8d83dd8d2cac7d2ce9d2dacc", "generation_seconds_batch": 15.066808530013077, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.294529438018799, "target_z": 0.4335936819801867, "ce": 5.383228778839111, "genuineness": 3.948779582977295, "identity": 0.0740705132484436, "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_CART_reg/38943f8ae18f8988be06e530ec4799dccf105fa0f9f222c3d418b283bda80542.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "38c751b989c32965d0649cc3b09e8b581275219c32422ece5b6591d0a73316fb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "907c7cddf49f2fb22dd888192a2ccfece4d596ef96bf7a2476e8f25369795e67", "generation_seconds_batch": 22.94136196898762, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9602538347244263, "target_z": 0.1894530937108859, "ce": 6.133810997009277, "genuineness": 4.2751336097717285, "identity": 0.4979882538318634, "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_CART_reg/38c751b989c32965d0649cc3b09e8b581275219c32422ece5b6591d0a73316fb.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "3952a422ab699831e1317a41f2d05ebd72f87b449ea1de92f0db55a217efc3cf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b3795ede841a206a289720b2b8cc0dfbdf27bb91b00cacf424ca22c68647f1d5", "generation_seconds_batch": 19.523576807929203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.684810757637024, "target_z": -0.011718736396037342, "ce": 6.078212261199951, "genuineness": 4.193545341491699, "identity": 0.5574241876602173, "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_CART_reg/3952a422ab699831e1317a41f2d05ebd72f87b449ea1de92f0db55a217efc3cf.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": "399e2d39a18de4dafd2e7db0dc737778c429ff9b10e4fe413db89c92c63c339e", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e43875b10233cff71950d8ab2763ac57b95b9b0727eb0334a23178df19ce5294", "generation_seconds_batch": 20.855196572025307, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3986707925796509, "target_z": -0.22070311547722615, "ce": 6.205295085906982, "genuineness": 4.457139015197754, "identity": 0.5898901224136353, "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_CART_reg/399e2d39a18de4dafd2e7db0dc737778c429ff9b10e4fe413db89c92c63c339e.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "39d6d2ffa3a5f5368e9d3913694f536b59908c32491c3f87c9603c4ad5ba3163", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "474a8057b526e3b02976b63a709708e66c5e4378e017e3f8cb26e986212bf16d", "generation_seconds_batch": 14.88069182599429, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.275657296180725, "target_z": -0.31054690628911413, "ce": 6.078212261199951, "genuineness": 4.3504462242126465, "identity": 0.43077218532562256, "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_CART_reg/39d6d2ffa3a5f5368e9d3913694f536b59908c32491c3f87c9603c4ad5ba3163.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "39ea432338ed73c5a9b96277181a2156ef6f28da13f6ce1aeb7ccbb016a6e38f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "370702df917fd2d29b744a9797bbb0d588bd98f6467d825fdbb970864c706598", "generation_seconds_batch": 16.57857537304517, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7651896476745605, "target_z": 0.7773437472792075, "ce": 5.383228778839111, "genuineness": 3.6663577556610107, "identity": 0.23306381702423096, "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_CART_reg/39ea432338ed73c5a9b96277181a2156ef6f28da13f6ce1aeb7ccbb016a6e38f.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "3a040b59445deedaf4513b04f94f8f9d5f565cad9017f0f12782a10dfd11da7e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "76b080c27b550cfde1aff62a3e2735bf03ba5831c63e4f9815ed55b0dc3e5a9c", "generation_seconds_batch": 21.913951030001044, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5056389570236206, "target_z": -0.14257814812673653, "ce": 6.181467056274414, "genuineness": 4.369274139404297, "identity": 0.6965949535369873, "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_CART_reg/3a040b59445deedaf4513b04f94f8f9d5f565cad9017f0f12782a10dfd11da7e.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "3a28ecb1309ed1ea382f6c6ae042f38bdafee0b1c2b3d9c72e57335f96c72d03", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b88c548b2d483a68cd0951833c7b61523a54be6b39d05ffe34ab49c0272c8928", "generation_seconds_batch": 14.985067666973919, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.203606605529785, "target_z": 0.36718753809109544, "ce": 5.824047088623047, "genuineness": 4.20609712600708, "identity": 0.5795135498046875, "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_CART_reg/3a28ecb1309ed1ea382f6c6ae042f38bdafee0b1c2b3d9c72e57335f96c72d03.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "3a59076dabb76c5a57bac8101fb0cd823f8ad7a454f88884e3fc3a48a18ebef5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "50e4e2063fe13f6e2b0d427433732deaec83fa9e7ac07bf3ab6ba8fd0201ed96", "generation_seconds_batch": 21.83409894502256, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8504586815834045, "target_z": -0.6210937690455477, "ce": 6.149696350097656, "genuineness": 4.507347106933594, "identity": 0.40472927689552307, "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": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/3a59076dabb76c5a57bac8101fb0cd823f8ad7a454f88884e3fc3a48a18ebef5.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "3b5b8a021561bd0d38669c3f8487f223cf81439c22776f3a4f0dc16a4d09acb4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b393460fbc0929285c7e3046df555c5bf2604f50fb01dcb790479ef0a73a3b9f", "generation_seconds_batch": 20.262918774038553, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6266467571258545, "target_z": -0.05419923201386359, "ce": 6.09409761428833, "genuineness": 4.250029563903809, "identity": 0.7254946231842041, "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_CART_reg/3b5b8a021561bd0d38669c3f8487f223cf81439c22776f3a4f0dc16a4d09acb4.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "3b67fe70f2beb8d9e15e8c71e1a9461fb8d79b08bb383fc6f062f712b141ebce", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ce9d1adb689aa090db210aee8d77f1dfc1b29fad477a7f2024dbe12b7823d590", "generation_seconds_batch": 19.898845841991715, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5404036045074463, "target_z": -0.11718753809109544, "ce": 6.268836498260498, "genuineness": 4.394378662109375, "identity": 0.4625677168369293, "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_CART_reg/3b67fe70f2beb8d9e15e8c71e1a9461fb8d79b08bb383fc6f062f712b141ebce.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "3c0b55529ab9be03c18cf6ac4cc485e06c533cab8f8d425e3037c40720abbd38", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4595d2e6352da02ee9ce8563b73a2a818b8dfaf5d20da3eb70e0f5550795e941", "generation_seconds_batch": 19.749617695924826, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.203606605529785, "target_z": 0.36718753809109544, "ce": 5.919358730316162, "genuineness": 4.149612903594971, "identity": 0.7688384056091309, "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_CART_reg/3c0b55529ab9be03c18cf6ac4cc485e06c533cab8f8d425e3037c40720abbd38.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "cf5474fb985b52d78021d4c754429563b11e2cf8184613db86bd55e58281a773", "generation_seconds_batch": 15.212075505987741, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6204770803451538, "target_z": -0.7890624836752448, "ce": 6.292664527893066, "genuineness": 4.149612903594971, "identity": 0.7418432235717773, "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_CART_reg/3c590f9302048ad206a88d3c2440f1028cc9eea1f1662250baf365be21a6ebfa.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "3c8f2a5d2ce03b84c7e7c3515ccdaa013945521cf7d34363a2d3e2035a03bd14", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "df11b91f59a9d6226ee19487d4d4c44e936a3365f8943ed1faf98360ead06ad4", "generation_seconds_batch": 19.66804720496293, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.104660987854004, "target_z": 0.294921895405944, "ce": 5.601652145385742, "genuineness": 4.124508857727051, "identity": -0.12348699569702148, "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_CART_reg/3c8f2a5d2ce03b84c7e7c3515ccdaa013945521cf7d34363a2d3e2035a03bd14.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "3c99f7da68dd5b25b6ee59ea18325e7c720dbb55325081c9839dc7f0dba032d3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "764063572eb1e9dbecf66d71c5fed29bb288c3986d9687123d25a3dc7ae42d62", "generation_seconds_batch": 18.882992491940968, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9844744205474854, "target_z": 0.9374999564673195, "ce": 5.5817952156066895, "genuineness": 3.760498285293579, "identity": 0.6148741245269775, "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_CART_reg/3c99f7da68dd5b25b6ee59ea18325e7c720dbb55325081c9839dc7f0dba032d3.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "3cb73ee605305277680f01ee453df006fc4b3b04b98c2ade81fdc2d70fbbaa9c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "43625b042ec631c327471b96c658b5323acf8cd886c5c33c5a402834812910a2", "generation_seconds_batch": 20.035419023944996, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.465373158454895, "target_z": -0.9023437472792075, "ce": 6.141753673553467, "genuineness": 4.632867813110352, "identity": 0.6600050926208496, "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_CART_reg/3cb73ee605305277680f01ee453df006fc4b3b04b98c2ade81fdc2d70fbbaa9c.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "3cd7ea3b418d8424ed4875a4c6fed13d72a6a0c2292014ca393a9bec7761a56c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "69d68952eefdbe7ccb6a04342ce327f27e5b82fd548c85e7bb4cb0750440c7c5", "generation_seconds_batch": 23.003980839974247, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.045828342437744, "target_z": 0.2519530501782054, "ce": 6.086154937744141, "genuineness": 4.030367851257324, "identity": 0.37959858775138855, "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_CART_reg/3cd7ea3b418d8424ed4875a4c6fed13d72a6a0c2292014ca393a9bec7761a56c.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "3ce64d3a9a151b76593aafa526462340c3e3556de9b0a4dd96ba1391eeac08f5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "df47212eb0ac3b02d593ddec8f9983eed1469c638af20e20c04a4cc24f99791c", "generation_seconds_batch": 19.29071022104472, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9201407432556152, "target_z": 0.16015620918811202, "ce": 6.086154937744141, "genuineness": 4.162164688110352, "identity": 0.3635420799255371, "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_CART_reg/3ce64d3a9a151b76593aafa526462340c3e3556de9b0a4dd96ba1391eeac08f5.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "3d105f79bfd32da7e0098aac711806db8e6e255f1e1510b7874356f2b42e1feb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d01dfe26b99136c60878b2c71b808f1ef371bbeb3b88cc2e91b2a73663ed11d9", "generation_seconds_batch": 19.7845632720273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8454158306121826, "target_z": 0.8359375163247552, "ce": 5.001980304718018, "genuineness": 3.6098732948303223, "identity": 0.19233939051628113, "hypothesis": "Heb ik er het kleine nouwtje gevonden? Het stond naast het oude raam! We kunnen er samen in kijken.", "wer": 0.16666666666666666, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/3d105f79bfd32da7e0098aac711806db8e6e255f1e1510b7874356f2b42e1feb.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "3d2c7b4bcaf003e05d28d0a9871b279846d5faf0c1260493d46be6693a1db716", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8718ad881654363b3e54fb4185a89f2aabb8c3e67ab1f8ba1f8380ff85a0fd78", "generation_seconds_batch": 22.9633475280134, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8611555099487305, "target_z": -0.6132812636039626, "ce": 6.507116317749023, "genuineness": 4.770940780639648, "identity": 0.4428741931915283, "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_CART_reg/3d2c7b4bcaf003e05d28d0a9871b279846d5faf0c1260493d46be6693a1db716.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "3dcca4edf95b1af1434b33a65fdbd144be0205a803f7df647d4baad717c7ea0c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "bd549f52d25ae6c804279c9c1dfe2f1d88b7331d47ca4811e5392a0434099b5a", "generation_seconds_batch": 19.758717976976186, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0724177360534668, "target_z": -0.4589843790811888, "ce": 6.221180438995361, "genuineness": 4.607763767242432, "identity": 0.6217935085296631, "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_CART_reg/3dcca4edf95b1af1434b33a65fdbd144be0205a803f7df647d4baad717c7ea0c.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": "3de445ea4a06faabe10863f9a60c1cbd39b3bb04efad0c015ea30185658896b2", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f2704ef891962c3d7f71d0a65c2690fe45097ce69e72dd91e41c50c281458fc8", "generation_seconds_batch": 20.448546694009565, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0536983013153076, "target_z": -0.47265625272079254, "ce": 6.443574905395508, "genuineness": 4.582659721374512, "identity": 0.625120997428894, "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_CART_reg/3de445ea4a06faabe10863f9a60c1cbd39b3bb04efad0c015ea30185658896b2.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "3e2a226c48a7b26a7e40f9239402e3fc98ac0693f68c144e1d769a32ed114500", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "38cb5f61ddf6ec10404038d2fb0c0edd616131b027b3bbd49fb273312bcf8932", "generation_seconds_batch": 18.936072392971255, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8718523383140564, "target_z": -0.6054687581623776, "ce": 6.403861999511719, "genuineness": 4.18099308013916, "identity": 0.40144139528274536, "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_CART_reg/3e2a226c48a7b26a7e40f9239402e3fc98ac0693f68c144e1d769a32ed114500.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "3eb7808d50cd104ce84f4ae336975d33aaf81fec887c5d7aa6de5d73369f6084", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ba7d027793e82d587758075be8b54e3b5d20334bf74189f3cf2ccac92d8f3038", "generation_seconds_batch": 23.823033366934396, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2383711338043213, "target_z": 0.3925780610613755, "ce": 6.1020402908325195, "genuineness": 4.2751336097717285, "identity": 0.6510487794876099, "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_CART_reg/3eb7808d50cd104ce84f4ae336975d33aaf81fec887c5d7aa6de5d73369f6084.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "3efa4705d6bdbfe3b1e55c0aee73d4bda0227e600309690ad8f518fe162120d8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "65f319eadd6e0848ca5bd0280ad0db9989c91e3242d224c12365998cd81011bb", "generation_seconds_batch": 21.64173847506754, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.59938907623291, "target_z": 0.6562501088317013, "ce": 5.681078910827637, "genuineness": 3.8797430992126465, "identity": 0.5730811953544617, "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_CART_reg/3efa4705d6bdbfe3b1e55c0aee73d4bda0227e600309690ad8f518fe162120d8.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "3f463d7a45fb55f726de460667eccb5ea9b8e886c42558ddcf71614cf9960c70", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8d43840185a0825cba2af0a37fef8223c8e3ee1453c8f0106bb6b9f297771492", "generation_seconds_batch": 21.24153371399734, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1098566055297852, "target_z": -0.4316406318019813, "ce": 6.229123115539551, "genuineness": 4.469690799713135, "identity": 0.33251452445983887, "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_CART_reg/3f463d7a45fb55f726de460667eccb5ea9b8e886c42558ddcf71614cf9960c70.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "3f99c57657d6f453a452070d0a2116b16f6378e91d5962eba054aabec1aef28f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0769dd66f538e4f3cc044d8bf9cc7349d601bf416224b4767954f4749d709d21", "generation_seconds_batch": 19.634755764971487, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9948657751083374, "target_z": -0.5156250108831701, "ce": 6.268836498260498, "genuineness": 4.444586753845215, "identity": 0.4996599555015564, "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_CART_reg/3f99c57657d6f453a452070d0a2116b16f6378e91d5962eba054aabec1aef28f.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": "3fa37db4420aea1906dca6dd6758817ab5d66879c9de298a6110bbe9de13203c", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "675de889f0acc3be564512be0b77019d5a81618cbf3b89952712a0a324fab6e5", "generation_seconds_batch": 21.10331853700336, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5297068357467651, "target_z": -0.125, "ce": 6.157639026641846, "genuineness": 4.2751336097717285, "identity": 0.7706931829452515, "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_CART_reg/3fa37db4420aea1906dca6dd6758817ab5d66879c9de298a6110bbe9de13203c.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": "3fe199c88fc0e1110763992ef61aa075b40f0519c3f28dd8078a388851bd8677", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The road turns left behind the bridge.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ae03a08fb9d69eff8a4956dd21bee369589bd8bdd925fcbf070c45bd57fb7e8f", "generation_seconds_batch": 16.964236039901152, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2596120834350586, "target_z": -0.32226564268515145, "ce": 6.3641486167907715, "genuineness": 4.080576419830322, "identity": 0.4936738908290863, "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_CART_reg/3fe199c88fc0e1110763992ef61aa075b40f0519c3f28dd8078a388851bd8677.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": "40222fa4179d4fb13bfcc144bdadb77f9e2f04ce5f0654d6bd2c6d5d065c49d4", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The keys are on the kitchen table.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "342c87f90a3d34905634617b6e13c87912cde72341d1d2aa881cfdf1f5e2da10", "generation_seconds_batch": 21.014885128941387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3906481266021729, "target_z": -0.22656252720792533, "ce": 6.213237762451172, "genuineness": 4.005263805389404, "identity": 0.4238072633743286, "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_CART_reg/40222fa4179d4fb13bfcc144bdadb77f9e2f04ce5f0654d6bd2c6d5d065c49d4.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "408d13ae90662bd343808f3976f42d7160f323e155e1d9ad61b4bbaac63153b0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c15996c8f677f75a9b16d34dc6f13ee3de95c0c63e862f4a25522eb12761dc24", "generation_seconds_batch": 19.865049802931026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.007052421569824, "target_z": 0.2236327778098952, "ce": 6.197352409362793, "genuineness": 4.250029563903809, "identity": 0.23079082369804382, "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_CART_reg/408d13ae90662bd343808f3976f42d7160f323e155e1d9ad61b4bbaac63153b0.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "40be6bc93dac58d1f7c125e21c18cdcafc6cd27005410462fa281d1ba61e857d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "326e2115d7a463902717d51ac5d6560fe29f0e578b73f614701854152bea9597", "generation_seconds_batch": 19.199448053957894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3505351543426514, "target_z": -0.2558593246653382, "ce": 6.038498878479004, "genuineness": 4.281409740447998, "identity": 0.3186235725879669, "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_CART_reg/40be6bc93dac58d1f7c125e21c18cdcafc6cd27005410462fa281d1ba61e857d.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "41e848b9b200bad2d059590f1a4c856ba95b530fb5b1368332db12dc937f40f5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5df1d74313b5e566335feef26ed290c92ab5c886232357f6934f9db4978e59d9", "generation_seconds_batch": 15.128914439002983, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1740374565124512, "target_z": -0.38476568621783197, "ce": 6.332377910614014, "genuineness": 4.287685871124268, "identity": 0.6066208481788635, "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_CART_reg/41e848b9b200bad2d059590f1a4c856ba95b530fb5b1368332db12dc937f40f5.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "41f070fc48282803083b211b1c94bad5614efd9738bc7f52034206ac051a989e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "11a8c5260692fde630c85dfef27955b77c13d8c680ea7e8530151a08f47295d6", "generation_seconds_batch": 19.788465181947686, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.866809368133545, "target_z": 0.8515624401425643, "ce": 5.514282703399658, "genuineness": 3.9927117824554443, "identity": 0.10484083741903305, "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_CART_reg/41f070fc48282803083b211b1c94bad5614efd9738bc7f52034206ac051a989e.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": "41f61640ee2945b9f4d7e3fa168054819b6a51ce36a5eaf4c38bbc12c0705707", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a4dc608ff2d437e52d29a090edc607f1c9bb1d2ba06c86101339f19235e4917f", "generation_seconds_batch": 21.095746322069317, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0563725233078003, "target_z": -0.47070311547722615, "ce": 6.221180438995361, "genuineness": 4.419482707977295, "identity": 0.7727405428886414, "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_CART_reg/41f61640ee2945b9f4d7e3fa168054819b6a51ce36a5eaf4c38bbc12c0705707.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "42560237196c7694c799696cca512834a5aa5447848b6ec5e596fae7c6dc3ae2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c9cd1a08dd1ce092ccc1e30152ecb1da7234cec2cadbb0b57a12e990436ca7a5", "generation_seconds_batch": 22.108001183951274, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4361096620559692, "target_z": -0.19335936819801866, "ce": 6.173524379730225, "genuineness": 4.20609712600708, "identity": 0.6343130469322205, "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_CART_reg/42560237196c7694c799696cca512834a5aa5447848b6ec5e596fae7c6dc3ae2.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "426e6001ba6dd0eb8ee3fe2927d4c7523def848c4d71322fb8e6e9cc714813f5", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e772ea9e4b07833ef45378c0a7aff671c3438b3f3d6914d5be3a82a283c152da", "generation_seconds_batch": 16.510753549053334, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4361096620559692, "target_z": -0.19335936819801866, "ce": 6.117925643920898, "genuineness": 4.645420074462891, "identity": 0.2849094271659851, "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_CART_reg/426e6001ba6dd0eb8ee3fe2927d4c7523def848c4d71322fb8e6e9cc714813f5.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "427c90dc43edb1cf15b0bd26e575979275cac7ffec7cc8b9813ef6e008e8481a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "947c769e2b653317a2af5b8ce871e255ec3b1a489082a749e19b2328c1836d29", "generation_seconds_batch": 19.844987511984073, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6500461101531982, "target_z": -0.03710934643167842, "ce": 6.165581703186035, "genuineness": 4.243753433227539, "identity": 0.4056457579135895, "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_CART_reg/427c90dc43edb1cf15b0bd26e575979275cac7ffec7cc8b9813ef6e008e8481a.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "4285073edc6dc4b13b29c0c9eab687b562c2abc0d8debfa009ee3e59f68a4988", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "618d7c53bd01030c43848f4eae8a539985f36b64029e45f8d3a8e29143943ada", "generation_seconds_batch": 19.427252536988817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.265113353729248, "target_z": 0.41210943349703943, "ce": 5.935244083404541, "genuineness": 4.093128204345703, "identity": 0.453085333108902, "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_CART_reg/4285073edc6dc4b13b29c0c9eab687b562c2abc0d8debfa009ee3e59f68a4988.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "42a57cdd978a21b0c8cab74aee4ad3a0f510906db2e7a31a9d176838f231dec8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1c6e06b775001763eb8f100e29f512a66928fa2dca8b2f4e0b57c20f99243ee2", "generation_seconds_batch": 19.13083522208035, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0832672119140625, "target_z": 0.27929679745741287, "ce": 6.062326908111572, "genuineness": 4.268857479095459, "identity": 0.5960349440574646, "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_CART_reg/42a57cdd978a21b0c8cab74aee4ad3a0f510906db2e7a31a9d176838f231dec8.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "42f19047654a80f78da89fef36e100e9ed499ad668a09b10b9efcfe5856f2ef8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c605d4c18d7bcbea15b92b47087e07cfd415279b495de81384acf3ef47c26613", "generation_seconds_batch": 15.02997711894568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.438936710357666, "target_z": 0.5390625707406058, "ce": 5.967014789581299, "genuineness": 4.2249250411987305, "identity": 0.32042449712753296, "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_CART_reg/42f19047654a80f78da89fef36e100e9ed499ad668a09b10b9efcfe5856f2ef8.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "5277dbf17d39ff344106e79523acf793a5bfc00adee5e8041fe8944b33b220af", "generation_seconds_batch": 14.931478124926798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.46002471446990967, "target_z": -0.9062500217663403, "ce": 6.173524379730225, "genuineness": 4.469690799713135, "identity": 0.650660514831543, "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_CART_reg/434f953004ee3838b31e4f505ba0cb9202b34ab37936979fff50edb64a6a996f.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "436bfcc812354b0fa3185dec4159b83a6695ae2d5ee9e083bce2d94d7129cde5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d7f227e4ce6c146b158d8efd9471b778c2ca3dc0abf8a1bddf906e885cb09605", "generation_seconds_batch": 15.030100949923508, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7812347412109375, "target_z": 0.7890623966098838, "ce": 5.700935363769531, "genuineness": 4.049196243286133, "identity": 0.06772056221961975, "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_CART_reg/436bfcc812354b0fa3185dec4159b83a6695ae2d5ee9e083bce2d94d7129cde5.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "43ae2284f1f629dfd7d48e10b12a5117bbae29b46ed63ba8ef9f1613397a9f45", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b8dbe064fdb2b068688e61b7a27234d7af8cc384b85a2a8b7cde822ec2898c75", "generation_seconds_batch": 20.03790273994673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4656786918640137, "target_z": 0.5585937690455477, "ce": 5.792276382446289, "genuineness": 3.8420867919921875, "identity": 0.4918193221092224, "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_CART_reg/43ae2284f1f629dfd7d48e10b12a5117bbae29b46ed63ba8ef9f1613397a9f45.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "c04ce90c96e4b376ec65338ecf56779c760a549f60bf3f2a7af6f5f43c2225cf", "generation_seconds_batch": 15.902026745025069, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.46002471446990967, "target_z": -0.9062500217663403, "ce": 6.213237762451172, "genuineness": 4.469690799713135, "identity": 0.4532012343406677, "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_CART_reg/43d771a5c055e2aea1186dc99f4cdef2b4c6bc8c6b2779ad35a29ca11c75a91c.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": "43e55c7b8e2f042de3a6086a12b4092d34d0fab77531b8c03e78395f9588c761", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "19c22eae1b3cb006b7b3b66b594aa38e75c2608eac1cb4317e97917b44a7a531", "generation_seconds_batch": 21.105636605992913, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5297068357467651, "target_z": -0.125, "ce": 5.935244083404541, "genuineness": 4.080576419830322, "identity": 0.5814477205276489, "hypothesis": "De sleutels liggen op de keukertafel. Ik keek in de lach 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_CART_reg/43e55c7b8e2f042de3a6086a12b4092d34d0fab77531b8c03e78395f9588c761.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "44a0c0c9eff0d01cc27cabe3dddb2d2ff16d8dc194f04384b528fead5cdcee91", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "28e7f075424c13ac9f4b0b36efb677176f8d7307d5b540e9b3e422cccdef3f76", "generation_seconds_batch": 18.941351162968203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3639061450958252, "target_z": -0.2460937255128672, "ce": 6.268836498260498, "genuineness": 4.607763767242432, "identity": 0.39962249994277954, "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_CART_reg/44a0c0c9eff0d01cc27cabe3dddb2d2ff16d8dc194f04384b528fead5cdcee91.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "44e475511b816a70b28edc80fcd4cb6796fb4189e42d54c490e40b6905e4778b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fbc001f7c1811005c56d891d1e9720441646c931e5d0ef334169fb91330574de", "generation_seconds_batch": 15.678358871024102, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.036468744277954, "target_z": 0.24511720042376453, "ce": 6.125868320465088, "genuineness": 4.137060642242432, "identity": 0.4296284019947052, "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_CART_reg/44e475511b816a70b28edc80fcd4cb6796fb4189e42d54c490e40b6905e4778b.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "44e70e1704ee3608db8f8c4bd06c056ff6fbeefc4d93113306df22b711dab78c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b3ea742862dddf0b39e7e48d59aeeb5f162cd5e66ba7d2154f70fdeb0b5af99e", "generation_seconds_batch": 15.02997711894568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0966384410858154, "target_z": 0.2890625707406058, "ce": 6.046441555023193, "genuineness": 4.174716949462891, "identity": 0.25641393661499023, "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_CART_reg/44e70e1704ee3608db8f8c4bd06c056ff6fbeefc4d93113306df22b711dab78c.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "776b5a5384fa5d28eab61cfa44a43f0096271aac55e2593cb5384696c3f7892f", "generation_seconds_batch": 16.11792687396519, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9306848645210266, "target_z": -0.5625, "ce": 6.23706579208374, "genuineness": 3.9550557136535645, "identity": 0.37860506772994995, "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_CART_reg/45f9ec8497c8691f9e18d794cf98f47c522bfab041b36df5accabf5c5c456412.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "460f3ceead2e1457b9257fb7604f24ab321d5895e1669a8065059c22d6af5688", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "eb02ea7e2c09e7e6900bfd3449041b2a5c1e270928a44e80c977f12b5e7ded31", "generation_seconds_batch": 22.143152975011617, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3827781677246094, "target_z": 0.4980467756910726, "ce": 5.831989288330078, "genuineness": 3.9801597595214844, "identity": 0.4360411465167999, "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_CART_reg/460f3ceead2e1457b9257fb7604f24ab321d5895e1669a8065059c22d6af5688.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "478f7da8e0d2252c68f20d1b63e1252be87dc0ea2543845ae5aca60439890282", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "aac946dfe32dea7e5743156b5adce1cca85cb125faf832e32271f551915abdec", "generation_seconds_batch": 15.419269283069298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9146396517753601, "target_z": -0.5742187363960374, "ce": 6.419747352600098, "genuineness": 4.268857479095459, "identity": 0.47550168633461, "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_CART_reg/478f7da8e0d2252c68f20d1b63e1252be87dc0ea2543845ae5aca60439890282.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "0d6ca33e82304b8a4cbc6ee4e2a069face2ac71ffad9059537e6dc271918af82", "generation_seconds_batch": 14.667770802974701, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.45467638969421387, "target_z": -0.910156209188112, "ce": 6.070269584655762, "genuineness": 4.281409740447998, "identity": 0.646193265914917, "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_CART_reg/47d00b404f769e82194cfb39f2be7a6bd3cc5bfaac4f38356b0fada2abdb9f4d.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "47e51932ad1fd4d03ae13046c33c5a8169c1c5264c8faaa847739fe6c9358935", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "aa7027e75a266091a9a61a8c1b5f73c082f7a24b667c91e8d7b33a4965a192e5", "generation_seconds_batch": 19.850642392993905, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9656022787094116, "target_z": 0.19335936819801866, "ce": 6.308549880981445, "genuineness": 4.425758361816406, "identity": 0.6242640614509583, "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_CART_reg/47e51932ad1fd4d03ae13046c33c5a8169c1c5264c8faaa847739fe6c9358935.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "4861bb0be3dc3a26aaf3c31d04588ea1193c66805486bbcd6021aa71a4257b20", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7beb857f8b63b4d53b61089aff83a48cb2c9f7b2415beea5520d5deb2ab74c3c", "generation_seconds_batch": 21.810891795088537, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9898228645324707, "target_z": 0.9414062309544523, "ce": 5.752562999725342, "genuineness": 4.017816066741943, "identity": 0.3918178677558899, "hypothesis": "The door's still open! I put your goat on the chair! Tell me when you are ready!", "wer": 0.16666666666666666, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/4861bb0be3dc3a26aaf3c31d04588ea1193c66805486bbcd6021aa71a4257b20.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "48815c8e7e60b8e93d2eec84c24368c683ff991c7a3b710f783334779f1c1900", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c2a08a0aa0fc3d227a28001e30c009d83de0a3385ebf64f6b3346a40bc55a078", "generation_seconds_batch": 21.0297713939799, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9416871070861816, "target_z": 0.9062499347009793, "ce": 5.824047088623047, "genuineness": 3.9173994064331055, "identity": 0.5964095592498779, "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_CART_reg/48815c8e7e60b8e93d2eec84c24368c683ff991c7a3b710f783334779f1c1900.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "48bc39859d0115bc44cae3b7146962faf392e802706019a03b7d9e241b4b14c0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "385d8417fd5afc1aaf83cea60dcb15b55398084213200dc750c0ce129e95e33d", "generation_seconds_batch": 20.084777973010205, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.358863353729248, "target_z": 1.2109376033901162, "ce": 5.375286102294922, "genuineness": 3.986435890197754, "identity": 0.2792060375213623, "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_CART_reg/48bc39859d0115bc44cae3b7146962faf392e802706019a03b7d9e241b4b14c0.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "48e342d33aa9acc8461a377490e469a294ddfb6ea9da1f8996a24a5a600ea0a0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "2362f053da82423f507308ae601386158d46bab5d7040e94cf7c999fff8009cc", "generation_seconds_batch": 16.57857537304517, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8400673866271973, "target_z": 0.8320312418376224, "ce": 5.839932441711426, "genuineness": 3.961331605911255, "identity": 0.20223762094974518, "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_CART_reg/48e342d33aa9acc8461a377490e469a294ddfb6ea9da1f8996a24a5a600ea0a0.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "490891827ca1f3f57acbd78d12c8bc3078029eb6cfd01cabc8159aa2868947cf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0588c9080d4d4ed1c537e0231e49e83970e9cfa40e70661c50c94132e377b037", "generation_seconds_batch": 19.956939787021838, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0216078758239746, "target_z": -0.49609372551286723, "ce": 6.24500846862793, "genuineness": 4.3881025314331055, "identity": 0.6241639852523804, "hypothesis": "The porte is encouraged. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.2222222222222222, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/490891827ca1f3f57acbd78d12c8bc3078029eb6cfd01cabc8159aa2868947cf.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "49b5b66c1913ccd5dcd45ef053f005accfce36b71e9936589280d13d4686a0b0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "cd6f2155b4b0e8188aa8da43f7a14360f63cb7cf34987dd04dc6c881487c2afa", "generation_seconds_batch": 20.95564452605322, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.018933653831482, "target_z": -0.4980468627564336, "ce": 6.133810997009277, "genuineness": 4.507347106933594, "identity": 0.5396059155464172, "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_CART_reg/49b5b66c1913ccd5dcd45ef053f005accfce36b71e9936589280d13d4686a0b0.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "5118ddfa2e8e2022d722abbd079e8cc84c088a1a817507d8691601c2d0fbeeaa", "generation_seconds_batch": 14.715727481991053, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6953548192977905, "target_z": -0.7343749891168299, "ce": 6.117925643920898, "genuineness": 4.369274139404297, "identity": 0.5492568016052246, "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_CART_reg/49d468ae6540cc7c1e5b0e6883c25e2b29599458c9de0a5398000eaca62677ab.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "49f13921ad3d8419f0f2a538d124371e314ddb320bc34d320742bdbbcab0c2d2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a3f7db19f5dc3c0a5a697e1cce4e2eb82fdbe48da4048d8379bf1bc87ab0184c", "generation_seconds_batch": 16.792897250968963, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.829217791557312, "target_z": 0.09374997823365974, "ce": 6.157639026641846, "genuineness": 4.306513786315918, "identity": 0.21635574102401733, "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_CART_reg/49f13921ad3d8419f0f2a538d124371e314ddb320bc34d320742bdbbcab0c2d2.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "4a4340ec08b65ade3689ef204e17d5e8beca377bb5150603d6e3ce9cfe2473b1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "da3a8a3ab15f57215fbb9b4b7418f0c627f5821d80c208c91cf51b590da354b3", "generation_seconds_batch": 19.94459084700793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0378057956695557, "target_z": 0.2460937255128672, "ce": 6.181467056274414, "genuineness": 4.218649387359619, "identity": 0.6298466324806213, "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_CART_reg/4a4340ec08b65ade3689ef204e17d5e8beca377bb5150603d6e3ce9cfe2473b1.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "4a5ae3a131ff6d85da77721661144cb4c3fa180353d95cf43e55491dc36787e7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ad3f21f9226134ada123491f50cd2065e7fc448ed0d3648a5d0cdc3fa233bfe9", "generation_seconds_batch": 15.279747102060355, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0540037155151367, "target_z": 0.9882811765386017, "ce": 5.9431867599487305, "genuineness": 4.2249250411987305, "identity": 0.4480840265750885, "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_CART_reg/4a5ae3a131ff6d85da77721661144cb4c3fa180353d95cf43e55491dc36787e7.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "4a61a1d8fe098554467bd3db78b7aaee9175026f3fd45f445aa373889417c928", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "60dacd350a406e4a2c3bb626e93d28ef0e4f139e678b9b98992f020c4b612f90", "generation_seconds_batch": 20.262918774038553, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.040480136871338, "target_z": 0.24804694982179462, "ce": 5.760505676269531, "genuineness": 3.8922953605651855, "identity": 0.5598021745681763, "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_CART_reg/4a61a1d8fe098554467bd3db78b7aaee9175026f3fd45f445aa373889417c928.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "4abeff955a4a2a269ca49d67a3c0d036fffedb1297b82018ec73cbd5a1514c17", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d80da5be188da793a1f8d0cba2126d705d27e49279cf963675dc9d510308e295", "generation_seconds_batch": 15.506361396051943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9335118532180786, "target_z": 0.169921895405944, "ce": 6.1020402908325195, "genuineness": 4.18726921081543, "identity": 0.31645214557647705, "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_CART_reg/4abeff955a4a2a269ca49d67a3c0d036fffedb1297b82018ec73cbd5a1514c17.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "4b04faa6c30eb72ba3ea2e6470fb291a8de5fffb87b0ac03a1910094d61b6973", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a078b59269c9735452ddb80eecc5d74092a54c139a08afcba332a62d4958b5b7", "generation_seconds_batch": 14.946655928972177, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.869330883026123, "target_z": 0.1230468627564336, "ce": 6.221180438995361, "genuineness": 4.494794845581055, "identity": 0.5973907709121704, "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_CART_reg/4b04faa6c30eb72ba3ea2e6470fb291a8de5fffb87b0ac03a1910094d61b6973.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "4b05a9d5e8e04991a92e490cfcfbbba0b944a37bed011a5b56a6e8039f58cdaa", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f0272acf8ed80bacfa903d82e5c54bedb99313b502506ea5013fdc8d3e58e3ab", "generation_seconds_batch": 20.389921457972378, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5671457052230835, "target_z": -0.09765625272079254, "ce": 6.038498878479004, "genuineness": 3.911123275756836, "identity": 0.467033714056015, "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_CART_reg/4b05a9d5e8e04991a92e490cfcfbbba0b944a37bed011a5b56a6e8039f58cdaa.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "4b2347b40a9b6b6866a6fe26abbc8661e889677015d48abc4d36d1fcb550d89b", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "844f149c8f81ec44206ddb2db745dc7227954873158962e906247891a9119dcb", "generation_seconds_batch": 15.343296076054685, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5805166959762573, "target_z": -0.08789065356832158, "ce": 6.268836498260498, "genuineness": 4.65797233581543, "identity": 0.6417668461799622, "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_CART_reg/4b2347b40a9b6b6866a6fe26abbc8661e889677015d48abc4d36d1fcb550d89b.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "4c292b5c898cfc3b7b3d05db88c3251947346e647ee251905c4a2b9caae99c65", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e2152eb0bdf3b0878ef97c9adb4992b1a995dc4debbddc4039f928930428c55d", "generation_seconds_batch": 18.84776872396469, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5056389570236206, "target_z": -0.14257814812673653, "ce": 5.728734970092773, "genuineness": 4.18099308013916, "identity": 0.5133243799209595, "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_CART_reg/4c292b5c898cfc3b7b3d05db88c3251947346e647ee251905c4a2b9caae99c65.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "4ccdd4270a14aa490bd109e0810acbfede5c49a59ce1a801802642cff6777d3e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3aeb3fd0e9f341bd6ee97dd835ebc5883f5320f8836bdb36e42f433ecacf4246", "generation_seconds_batch": 21.277858664980158, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8880503177642822, "target_z": 0.13671873639603735, "ce": 6.078212261199951, "genuineness": 4.293961524963379, "identity": 0.47177815437316895, "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_CART_reg/4ccdd4270a14aa490bd109e0810acbfede5c49a59ce1a801802642cff6777d3e.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "4e0194824ba6f1bd80ab319c3c1ecb38fbf823907bb22e0ed5d07ab71594f8d4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7ce57f5ceeadaaf4858364f6a50d8994e584fe4842ac576ba6418cbac81e2a60", "generation_seconds_batch": 21.463669580989517, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9000842571258545, "target_z": 0.1455078104594056, "ce": 6.022613525390625, "genuineness": 4.042920112609863, "identity": 0.2815718650817871, "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_CART_reg/4e0194824ba6f1bd80ab319c3c1ecb38fbf823907bb22e0ed5d07ab71594f8d4.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "4e1736544383b874903e291b088ded36e1dc709c6a5cfb1e160dc2ce4986388e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "73690322374a771105f7ee59b3d258c7d153ddeac6bb9161a3a09ba91de16e2f", "generation_seconds_batch": 14.810796473990194, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.843925952911377, "target_z": 0.1044921895405944, "ce": 5.538110733032227, "genuineness": 4.030367851257324, "identity": 0.4910222291946411, "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_CART_reg/4e1736544383b874903e291b088ded36e1dc709c6a5cfb1e160dc2ce4986388e.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "4ec067dac4ed8c5dd0354451cfd6ce764032543d232c97d595410fae1fa796ae", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "24e990b1137ab4927fb6d3057536af620109f13ef25e81eaff195dbe00379b5f", "generation_seconds_batch": 20.03790273994673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.104660987854004, "target_z": 0.294921895405944, "ce": 5.967014789581299, "genuineness": 3.8295347690582275, "identity": 0.5740565657615662, "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_CART_reg/4ec067dac4ed8c5dd0354451cfd6ce764032543d232c97d595410fae1fa796ae.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "502fa9b755d0e0eb2ba719e1781d3346e8a6294176f9b2cf030ca5b78ad936ca", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "69a1c233bd441fb2915a32f9dca907a6655178035b8ef02af0e1e1251a6d9165", "generation_seconds_batch": 21.18606163095683, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9898228645324707, "target_z": 0.9414062309544523, "ce": 5.80021858215332, "genuineness": 3.848362922668457, "identity": 0.33272653818130493, "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_CART_reg/502fa9b755d0e0eb2ba719e1781d3346e8a6294176f9b2cf030ca5b78ad936ca.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "52ef39b413a8ca5188cce13edcd10054773b3b5aba2ed411269f3b73cff7ad71", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d4d02ac064a6df129d151361d4923b7cefe7af1dd158effb3a0de3041abc99b8", "generation_seconds_batch": 15.676996757974848, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4695372581481934, "target_z": -0.16894528325148028, "ce": 5.95112943649292, "genuineness": 4.049196243286133, "identity": 0.6856145858764648, "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_CART_reg/52ef39b413a8ca5188cce13edcd10054773b3b5aba2ed411269f3b73cff7ad71.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "53249789a8d8d8927e1f6ccfb21ed68ddd8e4d71569bb1d8f68ce8a7d95e9506", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0a53b1d3b09ccf82e7fda17064b9e6e1feb92cf4d5361a470b98dffdb6d6b243", "generation_seconds_batch": 19.759780065971427, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.968123733997345, "target_z": -0.5351562527207925, "ce": 6.038498878479004, "genuineness": 4.482243061065674, "identity": 0.3812599182128906, "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_CART_reg/53249789a8d8d8927e1f6ccfb21ed68ddd8e4d71569bb1d8f68ce8a7d95e9506.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "53611716bb1ba7ffdcd30de302651d426b12f504bedc435b22802a3d6afcd60f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c81ed9a3f1f02c3acecda6c20775e5f17c0ef69ceb94e6e90673562512071eb3", "generation_seconds_batch": 21.322745061013848, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.727750778198242, "target_z": 0.75, "ce": 5.240260601043701, "genuineness": 3.9048471450805664, "identity": 0.4455641806125641, "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_CART_reg/53611716bb1ba7ffdcd30de302651d426b12f504bedc435b22802a3d6afcd60f.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "cd375820bb1b9686250b52e9b312de71b34d3042ff5a07da4067e4a5efb6f469", "generation_seconds_batch": 16.11792687396519, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8237166404724121, "target_z": -0.6406250108831701, "ce": 6.292664527893066, "genuineness": 4.017816066741943, "identity": 0.6134839057922363, "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_CART_reg/546203f143d2dbd908be3097fd76d41d76e93afd1487fa15c64a3d119dff5d49.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "55177b29bd3a56184f675f6ea99bac3176c9b4a8939f3c75186217467e1316ac", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3f6d80faac0c3a57cd11d884483da78f1911f1b1ec353c6aba72f7baecd19c74", "generation_seconds_batch": 19.489968045963906, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8011386394500732, "target_z": 0.07324216777425414, "ce": 6.300607204437256, "genuineness": 4.20609712600708, "identity": 0.4802854061126709, "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_CART_reg/55177b29bd3a56184f675f6ea99bac3176c9b4a8939f3c75186217467e1316ac.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "556161bc0f16bf25184c488977ec617713f768d031aad62f78075ec08bcf3c1a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1f2e9f92980d81908bb237a9b28ad4b54b326406cb172a273536e4195811c9e7", "generation_seconds_batch": 19.752091440954246, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5405564308166504, "target_z": 0.6132812636039626, "ce": 5.681078910827637, "genuineness": 4.093128204345703, "identity": 0.5976923704147339, "hypothesis": "Oui ! J'ai trouvé la petite boîte. Elle était près de 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_CART_reg/556161bc0f16bf25184c488977ec617713f768d031aad62f78075ec08bcf3c1a.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": "55c276467f73bc65e5284059e6b5379b3e427c39d320d1ba11f44c917d9fad05", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "566d46646aa9dc820f396f22b994c31cbe744444fae1d54fbbed40ab28b885b0", "generation_seconds_batch": 15.302610380924307, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.695507526397705, "target_z": -0.003906274487132784, "ce": 6.157639026641846, "genuineness": 4.193545341491699, "identity": 0.8146737813949585, "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_CART_reg/55c276467f73bc65e5284059e6b5379b3e427c39d320d1ba11f44c917d9fad05.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "56278676b7dc815c521bb5b39fe19bd77dccf5224ae32dabddc854f01fec7a90", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "281a1f092e016dcf7a7ee6feddd482a5c236ecfb7df3ef5f1f630b557ec069df", "generation_seconds_batch": 19.230843696044758, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8706680536270142, "target_z": 0.12402347491089731, "ce": 6.070269584655762, "genuineness": 4.18726921081543, "identity": 0.2539168894290924, "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_CART_reg/56278676b7dc815c521bb5b39fe19bd77dccf5224ae32dabddc854f01fec7a90.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": "56fd81c43341fe921a4da59502b0c2bfd986860fd28ca6a48c1fcb80c38aee2f", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d1d8d759ffd8b59aefcb94189d4b1bf5ee9f9d086872d2453c1edcc9756a2662", "generation_seconds_batch": 21.105636605992913, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8746793270111084, "target_z": 0.12695313724356638, "ce": 6.006728172302246, "genuineness": 4.21237325668335, "identity": 0.5641471147537231, "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_CART_reg/56fd81c43341fe921a4da59502b0c2bfd986860fd28ca6a48c1fcb80c38aee2f.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": "572af1de633976f4981bff16c8739395be287eb56b3d8d6c596fe1e1445a3bf3", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ef1e4109fd2cfc419efed3df019a5c89927b98d5cef0c877539fcb2830e7a4d1", "generation_seconds_batch": 20.536085210973397, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7060516476631165, "target_z": -0.7265624836752448, "ce": 6.332377910614014, "genuineness": 4.362998008728027, "identity": 0.44938674569129944, "hypothesis": "Dies Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/572af1de633976f4981bff16c8739395be287eb56b3d8d6c596fe1e1445a3bf3.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "57371e786ccded5a6b8490c9ca137bb640a652ffa12184515665fda05730e600", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "072e5b174c0fb50546a3d121cc818e34470a49c658ffd4f25695e051acb1703b", "generation_seconds_batch": 19.684041789965704, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3264672756195068, "target_z": -0.2734374727920747, "ce": 6.308549880981445, "genuineness": 4.356721878051758, "identity": 0.3886759579181671, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Mantel auf den Stuhl gelegt. Sag mir, wenn du bereit bist. Sissi.", "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_CART_reg/57371e786ccded5a6b8490c9ca137bb640a652ffa12184515665fda05730e600.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "5836b7f43cc2280ce4b03b664769ebd77970deab1cfd23ef36ef0b863a51c6f8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5937d49802c477dffc70824d6e2fdc8d23670f02d66233e6d62b91fe48d45645", "generation_seconds_batch": 21.583754868945107, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.631479263305664, "target_z": 0.6796874074930539, "ce": 6.038498878479004, "genuineness": 4.093128204345703, "identity": 0.3783896863460541, "hypothesis": "I found the little box! It was beside the old window. You can look inside together.", "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_CART_reg/5836b7f43cc2280ce4b03b664769ebd77970deab1cfd23ef36ef0b863a51c6f8.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "58c0b8d099e29bd4ba9b93eb45150f856f5363624cdb3c2e3678a70efec6e217", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "140d0ab0f1ad2178cd36696388e707b853270d1aad7419203d117398645e22dd", "generation_seconds_batch": 16.022334815002978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9214779138565063, "target_z": 0.16113282134257573, "ce": 5.927301406860352, "genuineness": 4.124508857727051, "identity": 0.6340712308883667, "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_CART_reg/58c0b8d099e29bd4ba9b93eb45150f856f5363624cdb3c2e3678a70efec6e217.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "58c1cc04f6cc0c8b5980540b8c92b164f3a0add0a981eb28396c413aa142b2bf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "df5b9c75212360e7af0d7412d129f45caab685c2ff715cd601229b4cb03ac0b9", "generation_seconds_batch": 22.17708530300297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.104660987854004, "target_z": 0.294921895405944, "ce": 5.808161735534668, "genuineness": 4.218649387359619, "identity": 0.4353180527687073, "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_CART_reg/58c1cc04f6cc0c8b5980540b8c92b164f3a0add0a981eb28396c413aa142b2bf.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "58fd282136245dbdd8d8ebbf91f7ccb0604d0c8a1841d68b6adac161b49ee656", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d5b2951e0982e770d5fc1e2ba9cef0da184730319006d9acd2184e5d10c24395", "generation_seconds_batch": 18.76072395790834, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.072570562362671, "target_z": 0.2714844226138693, "ce": 5.974957466125488, "genuineness": 4.218649387359619, "identity": 0.3948325514793396, "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_CART_reg/58fd282136245dbdd8d8ebbf91f7ccb0604d0c8a1841d68b6adac161b49ee656.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "595fb6516f515c432c90814393d55cf629fa008a901c4a1aa7fb9052984d2a3b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e0d7ecea164bf8a6448052f78fdabcfd0087fe99be068e41241b9ef62dcf1e20", "generation_seconds_batch": 19.0001315230038, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9335118532180786, "target_z": 0.169921895405944, "ce": 6.109982967376709, "genuineness": 4.482243061065674, "identity": 0.5090126395225525, "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_CART_reg/595fb6516f515c432c90814393d55cf629fa008a901c4a1aa7fb9052984d2a3b.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "59efbed3b2ff8535c098b4f31d841763210a8c90b253338749db05e73e8b6fe5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ba275098e4793b60adbadf7806cfccc298eadef8359e8765596332e859d4fa00", "generation_seconds_batch": 15.700251981033944, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9202935695648193, "target_z": 0.8906250108831701, "ce": 5.450741291046143, "genuineness": 3.8922953605651855, "identity": 0.5059508085250854, "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_CART_reg/59efbed3b2ff8535c098b4f31d841763210a8c90b253338749db05e73e8b6fe5.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "5a477c336dcd73566ab92da90fe3672cc66968fcccae2d22c80857c189e82ecc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "804dfa0bcc816c0e4b35de271c93ab80404895d6bc56cf4bd05eaa3f40d58300", "generation_seconds_batch": 23.149161288980395, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.567298412322998, "target_z": 0.6328124619089046, "ce": 5.303802013397217, "genuineness": 3.6287014484405518, "identity": 0.2574659287929535, "hypothesis": "La puerta sigue abierto. He dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/5a477c336dcd73566ab92da90fe3672cc66968fcccae2d22c80857c189e82ecc.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "5a8b219a2358942e4b732b8cd44d1fcaa50986942c9ec217cb5ca6afb74fddcb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7315ea7c520ec3c37e4f0a99a60c5e67e959fffeda155e872553796580c9b56e", "generation_seconds_batch": 15.520689091994427, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4842453002929688, "target_z": -0.15820315900990664, "ce": 6.070269584655762, "genuineness": 4.2374773025512695, "identity": 0.5450974106788635, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd, zeg maar dat niet je klaar bent.", "wer": 0.10526315789473684, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/5a8b219a2358942e4b732b8cd44d1fcaa50986942c9ec217cb5ca6afb74fddcb.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": "5b286f0284d134585cb826776cda660e699711dfb692bea585b910b3ba4dd8bf", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "908b49c14a6c48dc555d3659d8db17d0421134f52d446865ccf3f7e01dd95dda", "generation_seconds_batch": 20.227202686015517, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.35305655002593994, "target_z": -0.9843749891168299, "ce": 6.284721851348877, "genuineness": 4.432034492492676, "identity": 0.7877997756004333, "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_CART_reg/5b286f0284d134585cb826776cda660e699711dfb692bea585b910b3ba4dd8bf.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "5b2dd82610c30154538ff2a3fdd51244710a1a689ca583040d3588e7486965c3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1d40457166a0575a40b60655ee3b45e7666f4290c9b353b628cbd3be63985428", "generation_seconds_batch": 20.44804618495982, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7449803352355957, "target_z": 0.03222654685544294, "ce": 6.157639026641846, "genuineness": 4.287685871124268, "identity": 0.2788585424423218, "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_CART_reg/5b2dd82610c30154538ff2a3fdd51244710a1a689ca583040d3588e7486965c3.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "5b7b1dbdca990d8c24cde60379890c6662bb297965c3ced7e244f897302aad9e", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e4c24f04dad0dfbd961f5fa36be9a888cadbd19102608e7e515d89a1b815d870", "generation_seconds_batch": 15.725315557909198, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2729830741882324, "target_z": -0.3125000435326805, "ce": 6.133810997009277, "genuineness": 3.9299514293670654, "identity": 0.3186503052711487, "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_CART_reg/5b7b1dbdca990d8c24cde60379890c6662bb297965c3ced7e244f897302aad9e.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": "5ba3e3966ab77bc7a9d9c1bdca578d000d87c1dfb76d4ea0538c6b1104bd4fb8", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b8db683ffeb1ff2b05cc2034de9a7bc41f02644cb533e6b12f8fdf5fd146f9b7", "generation_seconds_batch": 20.873691476066597, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3264672756195068, "target_z": -0.2734374727920747, "ce": 6.157639026641846, "genuineness": 4.444586753845215, "identity": 0.5502455234527588, "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_CART_reg/5ba3e3966ab77bc7a9d9c1bdca578d000d87c1dfb76d4ea0538c6b1104bd4fb8.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "5c0358b13551e2e3ca237df35298c9365d9a70cf0839c93626675bde0c607368", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3edc25de7a8d8a5dbaeff9ba80d66c02ec559f5e92cd6c090f7cf93785b68b9a", "generation_seconds_batch": 18.98166206595488, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.695507526397705, "target_z": -0.003906274487132784, "ce": 6.157639026641846, "genuineness": 4.3127899169921875, "identity": 0.1558932662010193, "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_CART_reg/5c0358b13551e2e3ca237df35298c9365d9a70cf0839c93626675bde0c607368.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "5c5d30c9ed54e7edbbf74f0b20b5b7972dd03373ed85d240357c114608a0c673", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c633f38761b24bc9a5b77b8c8527cb9d97d61cc133615dfad39a258e34695431", "generation_seconds_batch": 21.510647344985045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.406693458557129, "target_z": -0.21484370374652698, "ce": 6.149696350097656, "genuineness": 4.394378662109375, "identity": 0.6548093557357788, "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_CART_reg/5c5d30c9ed54e7edbbf74f0b20b5b7972dd03373ed85d240357c114608a0c673.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "5ca650af20b180f474fa3534c912ff77f4e8439d3bdc210a9202e159f67b7ff4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5a6e4e1a6363d2ef5259a394b98d6cbf322433a44aee8007a2097d360134cf8e", "generation_seconds_batch": 21.18606163095683, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6421761512756348, "target_z": 0.6874999564673195, "ce": 5.736677646636963, "genuineness": 3.4215922355651855, "identity": 0.5931441783905029, "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_CART_reg/5ca650af20b180f474fa3534c912ff77f4e8439d3bdc210a9202e159f67b7ff4.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "5cd7c6f5b4490edeec4a0fa03742fbfe052e8377cd476f6819cb1d9fe1f205cc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "77203ccbefd543ab1c44ad8609a8147c83c21012201d372c859c6ad06a9d7db7", "generation_seconds_batch": 18.955261257011443, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3585577011108398, "target_z": -0.25, "ce": 6.197352409362793, "genuineness": 4.15588903427124, "identity": 0.4099572002887726, "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_CART_reg/5cd7c6f5b4490edeec4a0fa03742fbfe052e8377cd476f6819cb1d9fe1f205cc.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "5ce838c044968db07efddc0be55e1d3b3af840c4b8d01fb593b34b0239f4f657", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "08a6807a7ad7120c098abc2630cf94aac16f776680e3e9b6fbbfd86a3696aac4", "generation_seconds_batch": 15.343296076054685, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3425124883651733, "target_z": -0.2617187363960373, "ce": 6.221180438995361, "genuineness": 4.21237325668335, "identity": 0.6819629669189453, "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_CART_reg/5ce838c044968db07efddc0be55e1d3b3af840c4b8d01fb593b34b0239f4f657.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "5d9157b27053c7273c536ecec2415d733c95fd2f1ec7d75cb4ec3ef164096a4e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e1dbd7d586aa1778f379473df80868deb635f21936d58ff97628713bbf7dff51", "generation_seconds_batch": 21.00786956003867, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.898900032043457, "target_z": 0.875000087065361, "ce": 5.824047088623047, "genuineness": 4.18726921081543, "identity": 0.4800495207309723, "hypothesis": "Chitou! J'ai trouvé la petite boîte! Elle était près de la vieille fenêtre! Nous pouvons regarder ensemble ce qu'elle contient.", "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_CART_reg/5d9157b27053c7273c536ecec2415d733c95fd2f1ec7d75cb4ec3ef164096a4e.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "5deb9bbeb8330315a104e52d8e02b41dfd0b01a39d050b08a38b9faf336ac738", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 129, "seconds_generated": 10.32, "mp3_sha256": "a59def2a3c68f4ff9b555214881325d36a37295769172ff23b3e1b859d8cc626", "generation_seconds_batch": 15.470006876043044, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.171516180038452, "target_z": 0.34375006529902075, "ce": 4.741609573364258, "genuineness": 3.6663577556610107, "identity": 0.28055328130722046, "hypothesis": "¡La puerta sigue abierta! ¡He dejado tu abrigo sobre la silla! Dime cuándo estás listo.", "wer": 0.0, "duration_error_s": 0.120000000000001, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/5deb9bbeb8330315a104e52d8e02b41dfd0b01a39d050b08a38b9faf336ac738.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "5dfb748b08f2081ffefe52e4ac7bb4123197bea91931fd7f478098bb10b958be", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "28d33d96b871458556593db8e74296ed0ba09a9a7df1f8be8249a87491d467cc", "generation_seconds_batch": 21.24153371399734, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7864305973052979, "target_z": 0.0625000435326805, "ce": 6.086154937744141, "genuineness": 4.331617832183838, "identity": 0.3151317238807678, "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": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/5dfb748b08f2081ffefe52e4ac7bb4123197bea91931fd7f478098bb10b958be.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "e17045f2fb2e399f1f478fecfe500807869973b15c50b7fa7da77d08d8715c37", "generation_seconds_batch": 15.92157692706678, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.1605137586593628, "target_z": -1.125, "ce": 6.157639026641846, "genuineness": 4.432034492492676, "identity": 0.8423882126808167, "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_CART_reg/5e0af124ad7421917e897f2876d49b25e213a1bc87e01fe43f09b9c255c2a32f.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": "5eb2fda50807353538b8b188dce501605b74c7f3e734b0b5870839feb8351b7a", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ac156a33f34f31eff5e27a2384778742d0af5b06f99974afd643460d16f002e0", "generation_seconds_batch": 20.85437680897303, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4120417833328247, "target_z": -0.2109375163247552, "ce": 6.395919322967529, "genuineness": 4.030367851257324, "identity": 0.38905659317970276, "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_CART_reg/5eb2fda50807353538b8b188dce501605b74c7f3e734b0b5870839feb8351b7a.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "5ed600192d98b011d10285afc0067d2155dc354d9b77f7b07e67a4d7dbb3238d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8a13a8620cb3f7492677fe7c008ffbe17518da2be72602cbdddba78de8adae0f", "generation_seconds_batch": 20.814423169009387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7911103963851929, "target_z": 0.06591796840990094, "ce": 6.1020402908325195, "genuineness": 3.9173994064331055, "identity": 0.35194334387779236, "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_CART_reg/5ed600192d98b011d10285afc0067d2155dc354d9b77f7b07e67a4d7dbb3238d.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "5eea1cd7e0b6bf4721577d4693b95e26110d4ccac9e5a72d12a91e9ed123f041", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f77a39d0e235344a1ab8e9eb4dded3f6516e4b9740772569afa4204fcdde2153", "generation_seconds_batch": 14.985067666973919, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6100857257843018, "target_z": 0.6640624836752448, "ce": 5.935244083404541, "genuineness": 4.15588903427124, "identity": 0.5906296968460083, "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_CART_reg/5eea1cd7e0b6bf4721577d4693b95e26110d4ccac9e5a72d12a91e9ed123f041.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "5efb070f1a6d080b52a904c2a0a59dca00bd47ef9d2a1999ddbf53046f6493cb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "9fde96b9eeaf6e1381de4ce26ceaa90451204c9ec471da03fbc91130be02bad6", "generation_seconds_batch": 21.010634328005835, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.605921745300293, "target_z": -0.06933589328712136, "ce": 6.213237762451172, "genuineness": 4.381826400756836, "identity": 0.32651981711387634, "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_CART_reg/5efb070f1a6d080b52a904c2a0a59dca00bd47ef9d2a1999ddbf53046f6493cb.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "5f10bbcf5c8e930c7c8e2f3c73f31b1bb5eabe5359476f30de2abe6d7bebba97", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8fb15bbb9330665408a88ce3f64ed6e9f1b2e0f4de3beb26904ecf387726ec37", "generation_seconds_batch": 15.716723824036308, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7969745993614197, "target_z": -0.6601562527207925, "ce": 6.252951145172119, "genuineness": 4.413206577301025, "identity": 0.6683648228645325, "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_CART_reg/5f10bbcf5c8e930c7c8e2f3c73f31b1bb5eabe5359476f30de2abe6d7bebba97.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "5fd212cba34b5deea61bf51c43f94ed5a0dd56e1a5a0980c8e4a1807fff359b6", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8fd1f411f77699fc891ea5b5ec4ae7efe287e0da3d5947fbc678169253609195", "generation_seconds_batch": 19.8398582249647, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4442849159240723, "target_z": 0.5429686710970166, "ce": 5.700935363769531, "genuineness": 3.9425034523010254, "identity": 0.042274873703718185, "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_CART_reg/5fd212cba34b5deea61bf51c43f94ed5a0dd56e1a5a0980c8e4a1807fff359b6.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": "5fe28d4ef64ee82a3da6c014376a396a58f49d929b054ab0c7719a905c6cb27e", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d7cad4ac67267abf45c1b04a8f7c659dde5484c3e69d34969e68033572379564", "generation_seconds_batch": 14.993886140990071, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1954312324523926, "target_z": -0.36914058826930085, "ce": 6.23706579208374, "genuineness": 4.632867813110352, "identity": 0.6139698028564453, "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_CART_reg/5fe28d4ef64ee82a3da6c014376a396a58f49d929b054ab0c7719a905c6cb27e.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "5fefa95ad3b322e62002a5d1b63241749f0c1301c51ef1eeacb68e4b5ffe5e4a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "98c2a7584f77fcce5320cd166ebc6909ed37dc747481dccd969c0bf12bca5292", "generation_seconds_batch": 18.987682837992907, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.297203779220581, "target_z": 0.43554690628911413, "ce": 6.109982967376709, "genuineness": 4.15588903427124, "identity": 0.5905634164810181, "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_CART_reg/5fefa95ad3b322e62002a5d1b63241749f0c1301c51ef1eeacb68e4b5ffe5e4a.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "601145973e29931532cffc4afa2547ee56d0471090a41c4363a241478fb3be0f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a7798210c6ce8e8c06ba3986d4b015e819d9ddc033d4c7a3864dbe936fbec189", "generation_seconds_batch": 19.752091440954246, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7117056846618652, "target_z": 0.7382813506693237, "ce": 5.959072113037109, "genuineness": 4.256305694580078, "identity": 0.48753622174263, "hypothesis": "J'ai trouvé la petite boîte! It is près de son foncourt! It is près de la vieille fenêtre! We pouvant regarder ensemble what it contient!", "wer": 0.631578947368421, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/601145973e29931532cffc4afa2547ee56d0471090a41c4363a241478fb3be0f.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "60bd969e0f31863baf08d556a60777271319d50bed30e21429cc4a682f5ee593", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d12752ecbb52e316f91b875de40b70b3353bbb4197ac3cce740406aca77cf015", "generation_seconds_batch": 25.19867157097906, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1556236743927, "target_z": 1.0625000435326806, "ce": 5.268059730529785, "genuineness": 3.7667744159698486, "identity": 0.2747868597507477, "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_CART_reg/60bd969e0f31863baf08d556a60777271319d50bed30e21429cc4a682f5ee593.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "61146b6b1d266a38b2fb9cc2da259988a7975fa1da3ea3a17181b30b3b7ed88f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ba957f2d3438dbd1b4b58870b4a59f0a4abb3c76dc11fa2c6d6e06a8c97a7be0", "generation_seconds_batch": 19.91401848499663, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7931160926818848, "target_z": 0.06738284310891599, "ce": 6.260893821716309, "genuineness": 4.049196243286133, "identity": 0.2523908019065857, "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_CART_reg/61146b6b1d266a38b2fb9cc2da259988a7975fa1da3ea3a17181b30b3b7ed88f.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "616417ab180142bd28026b90ec75801478ac8b583b134aab49478d51a28291bd", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5f035ec949401c14451f95caeb11248ed0fe944bd8dbd24a539403cd8eda9ec9", "generation_seconds_batch": 22.108001183951274, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3946595191955566, "target_z": -0.2236327778098952, "ce": 6.157639026641846, "genuineness": 4.243753433227539, "identity": 0.5291821360588074, "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_CART_reg/616417ab180142bd28026b90ec75801478ac8b583b134aab49478d51a28291bd.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "623fafd2b62cb68b6f350ef5a9e02f54dd492ec4ba52f1d03d0c4fe21a25d39a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "2fefc04a217c57852da127a7610c8cc182c04bef7ccb2ed03c6d4cacbed94069", "generation_seconds_batch": 20.674201525049284, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.856112480163574, "target_z": 0.8437498911682987, "ce": 5.788304805755615, "genuineness": 3.7667744159698486, "identity": 0.26945632696151733, "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_CART_reg/623fafd2b62cb68b6f350ef5a9e02f54dd492ec4ba52f1d03d0c4fe21a25d39a.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "62af316624d01565c5ee7e8db2290065220f1b5041d5823442af43316c8e1ab9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3ae453032d1215617a80f76eea7c38a9f4c02897d186bbbcc94cc076980c55d9", "generation_seconds_batch": 25.025240330956876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5566015243530273, "target_z": 0.624999912934639, "ce": 5.621508598327637, "genuineness": 3.7542223930358887, "identity": 0.4878772497177124, "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_CART_reg/62af316624d01565c5ee7e8db2290065220f1b5041d5823442af43316c8e1ab9.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "6321d1528d78cdcd452fc6be824800f010ae7359ddec880fed433eb3c132c173", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3e8a90974f6a8d9d19ddce21cbc9d85c4f05187f2990aff1f2ce03af7dbb4383", "generation_seconds_batch": 20.275696104974486, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6192927360534668, "target_z": -0.05957029413465041, "ce": 5.863759994506836, "genuineness": 4.036643981933594, "identity": 0.48915204405784607, "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_CART_reg/6321d1528d78cdcd452fc6be824800f010ae7359ddec880fed433eb3c132c173.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "6359e9d738f2785a072fbd8f5e10ef2da8ab8861b50f8db954e9471a7d5f8a29", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5d062d0996a3246412c898be2797015b9fbbc5ff4b080dff0971b042ab059a52", "generation_seconds_batch": 19.684041789965704, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3852996826171875, "target_z": -0.2304688016950581, "ce": 6.23706579208374, "genuineness": 4.362998008728027, "identity": 0.38010361790657043, "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_CART_reg/6359e9d738f2785a072fbd8f5e10ef2da8ab8861b50f8db954e9471a7d5f8a29.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "636a0654ee4059d24ea7c18c6912344d0308ac26895c597b60067a0704ecff67", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7d2ba2ba03eadf06427a919fd536b78b913da7c7148f97717798d8b6d01a0a7d", "generation_seconds_batch": 20.400047193979844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.470874309539795, "target_z": -0.1679687581623776, "ce": 6.292664527893066, "genuineness": 4.394378662109375, "identity": 0.324617475271225, "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_CART_reg/636a0654ee4059d24ea7c18c6912344d0308ac26895c597b60067a0704ecff67.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "640d4e720c064062ff310cba2afc3e2166954bcca00509e34d5586b2e59f774d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "81b17a0a6884533fa72a10a01f7c5c9bc1caaa5f1bb34d7df6da0238a9f3753c", "generation_seconds_batch": 18.965767242014408, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9388601779937744, "target_z": 0.17382808282771575, "ce": 6.260893821716309, "genuineness": 4.331617832183838, "identity": 0.6144224405288696, "hypothesis": "He encontró dado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.11764705882352941, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/640d4e720c064062ff310cba2afc3e2166954bcca00509e34d5586b2e59f774d.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "6526f03221184a0b3afea0d00b12894583dbaa68aa8fa9cc48006c50378e99da", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2302b551d5c2a23d0e9342ba14fb96bf61c8360d55a6a1740753f37929a5f058", "generation_seconds_batch": 18.985369650064968, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7327936887741089, "target_z": -0.7070312418376224, "ce": 6.427689552307129, "genuineness": 4.344170093536377, "identity": 0.3930248022079468, "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_CART_reg/6526f03221184a0b3afea0d00b12894583dbaa68aa8fa9cc48006c50378e99da.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "65c3d6b6d58c7bfaf1548cf443c9b754bfadc27ce1129dd9f8411480b1dc8193", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b96faabde1f03531bfbec283e62ae579ab435a7759a5de7d2da28825ef32a2a4", "generation_seconds_batch": 20.262258651899174, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5056389570236206, "target_z": -0.14257814812673653, "ce": 5.8875885009765625, "genuineness": 4.074300289154053, "identity": 0.41285592317581177, "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_CART_reg/65c3d6b6d58c7bfaf1548cf443c9b754bfadc27ce1129dd9f8411480b1dc8193.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "65c99070b3c75a96a2cdaf66b84c0ea54d18fc234076a51fcf715050b0782d24", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4a01f7af698c248eda27c29181316ecd019fba73078e7f5f9cf3927a584448b6", "generation_seconds_batch": 21.218505374039523, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.695660352706909, "target_z": 0.7265625272079254, "ce": 5.756534099578857, "genuineness": 4.055472373962402, "identity": 0.6028856039047241, "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_CART_reg/65c99070b3c75a96a2cdaf66b84c0ea54d18fc234076a51fcf715050b0782d24.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "65d6c5d53b5ec11df4446d3acfad311064a61fb8251f0f497829fba9bf59f9ca", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "38d0f066583333e43be013ee537fbfcb068be5c6ee7f8c2a638e83afdd22004a", "generation_seconds_batch": 21.868402069085278, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6207826137542725, "target_z": 0.6718750326495104, "ce": 5.911416053771973, "genuineness": 4.061748504638672, "identity": 0.779220461845398, "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_CART_reg/65d6c5d53b5ec11df4446d3acfad311064a61fb8251f0f497829fba9bf59f9ca.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "662e735a34e20f3d43f888b4224e9d2a888aceffbca5037131988848a7b720c5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "79493055931cdd20975bbdea1a2265f0b7e715a7be53f3d09be5c99f6e3a2ae5", "generation_seconds_batch": 21.52729752101004, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8878976106643677, "target_z": -0.5937499782336597, "ce": 6.149696350097656, "genuineness": 4.482243061065674, "identity": 0.5375961065292358, "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_CART_reg/662e735a34e20f3d43f888b4224e9d2a888aceffbca5037131988848a7b720c5.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "666044dc7bac92140c688d949301808f268e786da6eed2de12f85f1b51fbb806", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0087abf62e94971f4b6a8b69e4d87ef70de099dabd5fa44dcf391d838b6d0106", "generation_seconds_batch": 15.022237317985855, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7758865356445312, "target_z": 0.785156296253473, "ce": 5.990842819213867, "genuineness": 3.9173994064331055, "identity": 0.44489991664886475, "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_CART_reg/666044dc7bac92140c688d949301808f268e786da6eed2de12f85f1b51fbb806.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "66b9fbbd001430ebeb181ca4a934bb820db1bebbc2e975dc37896b7fe1e7d3d2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "dbc92a0327eb93bcf3daf795986a2c3128ba607581166c5a4bbe88883de4aa23", "generation_seconds_batch": 27.004941086983308, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0779190063476562, "target_z": 0.2753906971010021, "ce": 5.998785495758057, "genuineness": 4.281409740447998, "identity": 0.5680793523788452, "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_CART_reg/66b9fbbd001430ebeb181ca4a934bb820db1bebbc2e975dc37896b7fe1e7d3d2.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "66f70530587514a18e98f4867adfefb921550cd3922312d5342dea3d1b514c7f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9141c6cdf74a9fc7db72f3d6bafbd39dc23916bd622d79bef83c0b60c59aefb9", "generation_seconds_batch": 20.256401565042324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9361860752105713, "target_z": 0.17187503264951037, "ce": 6.09409761428833, "genuineness": 4.15588903427124, "identity": 0.3754268288612366, "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_CART_reg/66f70530587514a18e98f4867adfefb921550cd3922312d5342dea3d1b514c7f.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": "66fccc3ba3303c65c478b6fc9fad68c68576cd2578a0222ad357c70a1e0cf12c", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c2a26d7804d1ad45c429228f9f02259f5b25e847d6b457ceaeedc8d0d4b53c1c", "generation_seconds_batch": 20.873691476066597, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3679174184799194, "target_z": -0.24316406318019812, "ce": 5.855817794799805, "genuineness": 3.973883628845215, "identity": 0.6805135607719421, "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_CART_reg/66fccc3ba3303c65c478b6fc9fad68c68576cd2578a0222ad357c70a1e0cf12c.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "6770dd6aa78e8b278f1017573505fa3a3d5d270d1ace098d288f833f7ff26a74", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "86ea060350bcbe45e5a0bbef23ffcea191b3afa21851efd826165c7abeaee381", "generation_seconds_batch": 18.996058237971738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.795790195465088, "target_z": 0.06933589328712136, "ce": 6.213237762451172, "genuineness": 4.268857479095459, "identity": 0.5009118318557739, "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_CART_reg/6770dd6aa78e8b278f1017573505fa3a3d5d270d1ace098d288f833f7ff26a74.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "679654e85172b41dfa2f0de963506168f6759766f9822df49930f343358714ad", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "efcb96efbdcc87ebb8fb2c899babdbac7e054e1754a057558fda75ebd04e17a5", "generation_seconds_batch": 21.126510394969955, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2570905685424805, "target_z": 0.40624993470097925, "ce": 5.629451274871826, "genuineness": 3.7667744159698486, "identity": 0.47750145196914673, "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_CART_reg/679654e85172b41dfa2f0de963506168f6759766f9822df49930f343358714ad.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "67b7c895a97b65a4e21adba8ecd97da989c72823991cb59d2d57154f123c152b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ed9e572a770a6e2b874b6c0d1bd51a4eaaba43a8388fb4263406dbea8a45b5f4", "generation_seconds_batch": 20.031144544947892, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.570488452911377, "target_z": -0.09521485293267479, "ce": 5.974957466125488, "genuineness": 4.243753433227539, "identity": 0.7028353214263916, "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_CART_reg/67b7c895a97b65a4e21adba8ecd97da989c72823991cb59d2d57154f123c152b.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "684ae037e624205785ec4191b9d97d9b2e884722d074f6d7344c16472eb15b83", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "40d1f09e07895df4d3a0011ab7a9618747a2048ffd0944f3d7161e599c6003fd", "generation_seconds_batch": 15.066808530013077, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.2732887268066406, "target_z": 1.1484375598574357, "ce": 5.367343425750732, "genuineness": 3.6789097785949707, "identity": -0.10667965561151505, "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_CART_reg/684ae037e624205785ec4191b9d97d9b2e884722d074f6d7344c16472eb15b83.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "685bd93e651b1f47a5993504141b47e904a2246cf102f7b751a4ff6d45ebc38a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "59e99a773c53c710a446d93cd8f02ed66976f22b303aa47e02fd18a1054b76c8", "generation_seconds_batch": 20.97785390005447, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.64752459526062, "target_z": 0.6914062309544523, "ce": 5.816103935241699, "genuineness": 3.9425034523010254, "identity": 0.4464634358882904, "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_CART_reg/685bd93e651b1f47a5993504141b47e904a2246cf102f7b751a4ff6d45ebc38a.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "68727310ca258824f833d30d6880cb096d619c069d153d10a874160c760ca815", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "dd4e50e31aaf34c193d81e5b58b9ae4f1ad403e1c0c216cb56f8ef77413472e6", "generation_seconds_batch": 19.71047442907002, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5886921882629395, "target_z": 0.6484375598574357, "ce": 5.323658466339111, "genuineness": 3.8420867919921875, "identity": 0.12355153262615204, "hypothesis": "Goed, ik heb het kleine douche 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_CART_reg/68727310ca258824f833d30d6880cb096d619c069d153d10a874160c760ca815.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "68af55b74ebdb0d467e0dae36d5545923b90750029bf35a2fad3ec6762243fef", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2a0c743eb5bed7cf99370e57eb6d084cab066cfbc1111c38818e186f1c67531d", "generation_seconds_batch": 17.28488218097482, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7329463958740234, "target_z": 0.023437472792074684, "ce": 6.308549880981445, "genuineness": 4.243753433227539, "identity": 0.6294747591018677, "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_CART_reg/68af55b74ebdb0d467e0dae36d5545923b90750029bf35a2fad3ec6762243fef.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "6944aca624b20f5ec226ff39094a88d4b3ffef4a619e8ee67f5ba267000dfde0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "320ff2565e6f4f1b5c93e11d20192970aa735be4baa351e483107f639f2ce458", "generation_seconds_batch": 22.299648517975584, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.262439012527466, "target_z": 0.410156209188112, "ce": 6.117925643920898, "genuineness": 4.2374773025512695, "identity": 0.5746716856956482, "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_CART_reg/6944aca624b20f5ec226ff39094a88d4b3ffef4a619e8ee67f5ba267000dfde0.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "695100f41b69d383fff0e52034528cacb154646edd2c6af604b8101b6d467911", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "938bcc3f0122f6a0c6b2387a36088398b39f747fdfdf246b36068b3cd778204c", "generation_seconds_batch": 21.86206905997824, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.278484344482422, "target_z": 0.4218750326495104, "ce": 5.919358730316162, "genuineness": 4.068024158477783, "identity": 0.23438724875450134, "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_CART_reg/695100f41b69d383fff0e52034528cacb154646edd2c6af604b8101b6d467911.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "695e4d9955de55c4838ee918e0274853d553f6d69ccd834dc136c215e4a67dff", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7a2b5c5ffedfe57f6df27cb3f5b5a1014e313606d50b083a2c95b84ece3a10a8", "generation_seconds_batch": 14.943225278984755, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.416053056716919, "target_z": -0.2080078539920861, "ce": 6.205295085906982, "genuineness": 4.168440818786621, "identity": 0.8182791471481323, "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_CART_reg/695e4d9955de55c4838ee918e0274853d553f6d69ccd834dc136c215e4a67dff.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "69c3695e2a1bd694380f9af3551b7a762012b848488396d3dd572879618a6f31", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "96936f17e59529f2a78b576517d8a666141ce3c2debc7b1d0e40952545dc7dbf", "generation_seconds_batch": 19.739560261019506, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.265113353729248, "target_z": 0.41210943349703943, "ce": 5.847874641418457, "genuineness": 3.9927117824554443, "identity": 0.6434590816497803, "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_CART_reg/69c3695e2a1bd694380f9af3551b7a762012b848488396d3dd572879618a6f31.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "6a21c70292708f8975a3b5e3998371bc56f82b6b050b9f0afe270f4156b78f7e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3f33bed01c8f1de060f4b32749028fcc367eb3080b777bd79942e24b4178b2a8", "generation_seconds_batch": 20.45752632303629, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8935513496398926, "target_z": 0.8710936384475062, "ce": 5.246217727661133, "genuineness": 4.080576419830322, "identity": 0.6861514449119568, "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_CART_reg/6a21c70292708f8975a3b5e3998371bc56f82b6b050b9f0afe270f4156b78f7e.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "ee9695d0c1706623855a2d4bf295f7512a3e8cb5818727b3717d4baefa0c447a", "generation_seconds_batch": 16.05141171696596, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5562961101531982, "target_z": -0.8359375163247552, "ce": 6.300607204437256, "genuineness": 3.9048471450805664, "identity": 0.4322202503681183, "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_CART_reg/6a61bc20d43b3e97640d8dd45682f7055c5dbaa5aa987cdb55f8a7d7a432a2c2.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "6ac36fe388273bf7b8f17b715eae315525aac7d0b35b5f607fbd2ee9645d09e3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c8dde57dcab24f047b0a4b1daf89ffdb34e21a7440f717f96eac71e838b022c5", "generation_seconds_batch": 18.93665307399351, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6032475233078003, "target_z": -0.07128903053068775, "ce": 6.38003396987915, "genuineness": 3.8797430992126465, "identity": 0.4449467957019806, "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_CART_reg/6ac36fe388273bf7b8f17b715eae315525aac7d0b35b5f607fbd2ee9645d09e3.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "6accb6862b3ffb822103b16bafea8a24170a466b768b17900b5b3aa6e76ab9bc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e02fc20b9fe1f3140ca865342358cf464384af19f76411b00ee6ae3dce115ebb", "generation_seconds_batch": 25.28098876203876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9656022787094116, "target_z": 0.19335936819801866, "ce": 5.927301406860352, "genuineness": 4.030367851257324, "identity": 0.28858619928359985, "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_CART_reg/6accb6862b3ffb822103b16bafea8a24170a466b768b17900b5b3aa6e76ab9bc.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "6afb6bbdd9c73c7bd995d3db5b86acfedeb5d197316e9d8ea328fca4e8e4a1ed", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "af306a3cf9b0b8c36194c18ea531c16ea4b7c6cea186da74991615f326e18a97", "generation_seconds_batch": 21.15716633806005, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.738447666168213, "target_z": 0.7578125489742655, "ce": 5.748591423034668, "genuineness": 4.118232727050781, "identity": 0.41877543926239014, "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_CART_reg/6afb6bbdd9c73c7bd995d3db5b86acfedeb5d197316e9d8ea328fca4e8e4a1ed.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": "6b1a03c4dcdf78369b5b07d74a0ed856e46013284555840476d15070a9132b1a", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9fbc986ac0c23a8d72ef2f1edd5603a5b0ca03e50087650b93e9371c909e7b99", "generation_seconds_batch": 16.157840647967532, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4361096620559692, "target_z": -0.19335936819801866, "ce": 6.38003396987915, "genuineness": 4.557555198669434, "identity": 0.5324398875236511, "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_CART_reg/6b1a03c4dcdf78369b5b07d74a0ed856e46013284555840476d15070a9132b1a.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "6b2e60f7e9229ec593af46041de3ec8007641cdf9d0b94a8786f194f3cbe9e04", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a9ab769b4e221423a38b6929aac6e6d960c06859d3c3124b3386478dd9ae7b6c", "generation_seconds_batch": 21.339720066054724, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7115527391433716, "target_z": 0.007812461908904557, "ce": 6.09409761428833, "genuineness": 4.287685871124268, "identity": 0.5538626909255981, "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_CART_reg/6b2e60f7e9229ec593af46041de3ec8007641cdf9d0b94a8786f194f3cbe9e04.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "6b4b1c4faba5a0c45ba49c5927cb1ce53692e402991bbcf51766ae1b6a384224", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1549f14a33f260f16556b9f9170d2c38b575fa511bbe1393c2407916160d05ce", "generation_seconds_batch": 19.518725777044892, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3211188316345215, "target_z": -0.27734374727920746, "ce": 6.292664527893066, "genuineness": 4.369274139404297, "identity": 0.3586919605731964, "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_CART_reg/6b4b1c4faba5a0c45ba49c5927cb1ce53692e402991bbcf51766ae1b6a384224.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "49cab45c2f24ba3e7364a714939461118676c901aab344527f932f28aa78f0a0", "generation_seconds_batch": 15.902026745025069, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2810057401657104, "target_z": -0.3066406318019813, "ce": 6.332377910614014, "genuineness": 4.607763767242432, "identity": 0.7080582976341248, "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_CART_reg/6b973c750e3616c70f28663c29496a0c74bea4a87dff4bad65cadc5394b99131.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "6c0d68a5306c1052a85bfd95485c2a73e230f692f56ea87591b463e4a3dc6201", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "dcc6217d8b3fac466be1c5364fdbcf96edca56e04216bd99eec9a4c257cc6990", "generation_seconds_batch": 19.58179482491687, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1499696969985962, "target_z": -0.40234374727920746, "ce": 6.395919322967529, "genuineness": 4.18099308013916, "identity": 0.3611070513725281, "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_CART_reg/6c0d68a5306c1052a85bfd95485c2a73e230f692f56ea87591b463e4a3dc6201.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "6c733320fe60427db02fb1f7bea3b0cf378c724e4af7e7c26f7dc2f73d5c7221", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "653e1d09de88276725f115c8c3111fde74f8049c89751e36568c0ef8933bfeca", "generation_seconds_batch": 14.796144105028361, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6754510402679443, "target_z": -0.01855467321583921, "ce": 5.935244083404541, "genuineness": 3.923675537109375, "identity": 0.1866530328989029, "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_CART_reg/6c733320fe60427db02fb1f7bea3b0cf378c724e4af7e7c26f7dc2f73d5c7221.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "6c8c5492079933f631a80c0b0d76de619e8b0ff02954cee0fafb61947f96e7d2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a6e2f88aed4cf7aa82f0f583478768e6c09897c3d6cfc244d7fa9677a15b81b5", "generation_seconds_batch": 19.627846773015335, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9602538347244263, "target_z": 0.1894530937108859, "ce": 6.149696350097656, "genuineness": 4.319066047668457, "identity": 0.45244187116622925, "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_CART_reg/6c8c5492079933f631a80c0b0d76de619e8b0ff02954cee0fafb61947f96e7d2.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": "6d2f6a1a5162dc356438b4b1de7e20b3dec3609bbbfa333442ff159450e377d0", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dfc501725dc30fd741d76c1409caebf4cb98028ca11f47c11bfec0da5b21b8db", "generation_seconds_batch": 14.846530252019875, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4053562879562378, "target_z": -0.21582031590099066, "ce": 6.078212261199951, "genuineness": 4.432034492492676, "identity": 0.6726483106613159, "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_CART_reg/6d2f6a1a5162dc356438b4b1de7e20b3dec3609bbbfa333442ff159450e377d0.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "6dd790054d8a816b46eabdf27609886857efb57e1ece02511ca0c46bcb287c08", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "686c7ef89d19c9a6386820d3717d0b9998eaed171da54d165cc429afda917104", "generation_seconds_batch": 21.40882449503988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9788205623626709, "target_z": -0.5273437472792075, "ce": 6.046441555023193, "genuineness": 4.105680465698242, "identity": 0.3993273377418518, "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_CART_reg/6dd790054d8a816b46eabdf27609886857efb57e1ece02511ca0c46bcb287c08.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "6e0f60b53ba17a7e566dd6db30815996ff4ab688ca44e653f1ddfc56d8fc1bee", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3b7271efd365bb924fb950cc8f06ed02365ecb57636b8b88955d1696444e2bd1", "generation_seconds_batch": 20.503277458949015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9737777709960938, "target_z": 0.929687581623776, "ce": 5.154876708984375, "genuineness": 3.4969046115875244, "identity": 0.32761651277542114, "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_CART_reg/6e0f60b53ba17a7e566dd6db30815996ff4ab688ca44e653f1ddfc56d8fc1bee.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "6eeaaccf145a35bfd1faf19a027af3c31cec009b4b46b7a0d41d160de10db5c3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "44786f0218cc182e46661723274c3a8c7a3aaba56a5e46f8f9b32f9b940a5308", "generation_seconds_batch": 24.66216942097526, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4603302478790283, "target_z": 0.5546874945584149, "ce": 6.117925643920898, "genuineness": 3.9425034523010254, "identity": 0.47721001505851746, "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_CART_reg/6eeaaccf145a35bfd1faf19a027af3c31cec009b4b46b7a0d41d160de10db5c3.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "6f0353c846f4ddb6ac16b620630acbfd6651e9436bb7aff07d013f2dae59a3e7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d61532dd9290671ba150bd44b10764ab9b7d5ac4f07283e5c64c3a2ad7ecb9cf", "generation_seconds_batch": 15.520689091994427, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.275657296180725, "target_z": -0.31054690628911413, "ce": 6.054384231567383, "genuineness": 4.356721878051758, "identity": 0.10015404224395752, "hypothesis": "De deur staan nog open. Ik heb je jaselde op de stoel gelegd. Zeg maar 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/manual-comparison-v1/audio/vn_S_CART_reg/6f0353c846f4ddb6ac16b620630acbfd6651e9436bb7aff07d013f2dae59a3e7.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "6f1c6a8e0b149f88adb25101b788f52360424abbbfb707262e533b9ce7f3bb5c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "468d7b96010550c3ab84c773ca2abadab27aa7b7bd1fa3c1ed5d17459172c0f4", "generation_seconds_batch": 19.979299903963692, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.551100492477417, "target_z": -0.10937498911682987, "ce": 6.157639026641846, "genuineness": 4.413206577301025, "identity": 0.4715679883956909, "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_CART_reg/6f1c6a8e0b149f88adb25101b788f52360424abbbfb707262e533b9ce7f3bb5c.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "6f7804dd4aad93517dfaff2cef991610200eb0fbfb455c27a756ccad7f9326ff", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c69dfcfa8e0679eb7d3fd2c95761e12256c90198508796ac07187de57bbc7d8b", "generation_seconds_batch": 23.149161288980395, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.230501413345337, "target_z": 1.1171875380910954, "ce": 5.5817952156066895, "genuineness": 3.948779582977295, "identity": 0.38466981053352356, "hypothesis": "¡La puerta! ¡Sí que ha abierto! ¡He dejado tu abrigo sobre la silla! Dime cuándo estás listo.", "wer": 0.26666666666666666, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/6f7804dd4aad93517dfaff2cef991610200eb0fbfb455c27a756ccad7f9326ff.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "7042c2af0a910844b4da9810037cb0e1ce60895b20a8d2db18dbcd69e3359472", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "44139f1b429495c945fd8dc42eed617784482ef09ffb78a197d1137fff911b04", "generation_seconds_batch": 16.328598621999845, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1501224040985107, "target_z": 0.3281249673504896, "ce": 5.752562999725342, "genuineness": 4.193545341491699, "identity": 0.189274862408638, "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_CART_reg/7042c2af0a910844b4da9810037cb0e1ce60895b20a8d2db18dbcd69e3359472.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "717da8bddf9c4cddac6ef9afccd2d8b094e51fe5448f998f03a73ea0c41a8396", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3b4b248661f5e48d938263088fb451bc22ac5676d21eb7763ea301f98a9ab78f", "generation_seconds_batch": 21.476005648029968, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.64752459526062, "target_z": 0.6914062309544523, "ce": 6.141753673553467, "genuineness": 4.293961524963379, "identity": 0.64070725440979, "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_CART_reg/717da8bddf9c4cddac6ef9afccd2d8b094e51fe5448f998f03a73ea0c41a8396.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": "71c8883ac1d379f05cf266a4c7042ba808637a2984fe50411fbf9d4f050d070b", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Someone moved the little blue chair.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2ba017621fdc54c2151153f32b64c132fa9617545647b0c497bf893b51fa06a5", "generation_seconds_batch": 21.231656990014017, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.029630422592163, "target_z": -0.49023440084752906, "ce": 6.38003396987915, "genuineness": 4.457139015197754, "identity": 0.4787820279598236, "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_CART_reg/71c8883ac1d379f05cf266a4c7042ba808637a2984fe50411fbf9d4f050d070b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "71f0917541ad7172c84254188cc299c64fa34fcfc9860aeb8c361d54eb5d295e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "61b1f9efeacd1225ca99f06c22efbda3d966cef16dc05ff5849d9a69561bae2c", "generation_seconds_batch": 19.979299903963692, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.497463583946228, "target_z": -0.8789062744871328, "ce": 6.252951145172119, "genuineness": 4.65797233581543, "identity": 0.6997151970863342, "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_CART_reg/71f0917541ad7172c84254188cc299c64fa34fcfc9860aeb8c361d54eb5d295e.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "71f556e895ed8b47d32052b777f950e1e68f287815d526d22d4f1eb26a99bd02", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e34363bcebcaaa6fe4ab8db100b7517254e2920fc42c79870740f527f11c982c", "generation_seconds_batch": 19.898845841991715, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0418171882629395, "target_z": 0.2490234749108973, "ce": 6.173524379730225, "genuineness": 4.137060642242432, "identity": 0.40317460894584656, "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_CART_reg/71f556e895ed8b47d32052b777f950e1e68f287815d526d22d4f1eb26a99bd02.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "72302b3e155e76f68e98b5d832599ede0c04a495761e5bbf9237e90a82ffc559", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "39bc7184b95040c07566598d82b67007a90447435a7468e628283ea38641b216", "generation_seconds_batch": 20.43732152704615, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2489153146743774, "target_z": -0.330078104594056, "ce": 6.403861999511719, "genuineness": 4.300237655639648, "identity": 0.389604777097702, "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_CART_reg/72302b3e155e76f68e98b5d832599ede0c04a495761e5bbf9237e90a82ffc559.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "73348f86282b756d11792796b9bebdd7224ee5a89687d08fd562edc74d9ab5e0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fcb6e8755d62265704d305d77045ad8de9484a11d23ff773bb9045eb7caeec8d", "generation_seconds_batch": 15.591984582017176, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7491445541381836, "target_z": 0.7656250979485312, "ce": 5.689021587371826, "genuineness": 3.722842216491699, "identity": 0.5211837291717529, "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_CART_reg/73348f86282b756d11792796b9bebdd7224ee5a89687d08fd562edc74d9ab5e0.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "7343e54cef946f2c5eb8f85446a9a2cc1cb2e9c1bc61e7e8f5ae7ef548147aa0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4a42a2c224f155062d2f5e1d281b0e572920cd9b85782349c76cde41979b5925", "generation_seconds_batch": 19.230843696044758, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6580686569213867, "target_z": -0.03125002176634025, "ce": 6.300607204437256, "genuineness": 4.532451152801514, "identity": 0.6694679856300354, "hypothesis": "Ik heb het... Ik 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_CART_reg/7343e54cef946f2c5eb8f85446a9a2cc1cb2e9c1bc61e7e8f5ae7ef548147aa0.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "73adb07de61ff874c17b4de356bfc2093b6ea4de7ca020057e24acb33f8a3062", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "bd4a9231e06d4eafda0f1aa19c5617e38754c8a553b1ba96c24047a0463edc9e", "generation_seconds_batch": 20.262258651899174, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.091137170791626, "target_z": -0.44531250544158507, "ce": 5.974957466125488, "genuineness": 4.3127899169921875, "identity": 0.5056921243667603, "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_CART_reg/73adb07de61ff874c17b4de356bfc2093b6ea4de7ca020057e24acb33f8a3062.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "73f50f2fbd96c4e0351506d2ee8bc81004ac16c956332d2064c9c55614d5ea80", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "94a4f01c726f3bf53cd402b9d4e5d463a7f4c652805c24013bebb8b28aaa3f54", "generation_seconds_batch": 20.653390570892952, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.525695562362671, "target_z": -0.1279296623326691, "ce": 6.038498878479004, "genuineness": 4.080576419830322, "identity": 0.6751227378845215, "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_CART_reg/73f50f2fbd96c4e0351506d2ee8bc81004ac16c956332d2064c9c55614d5ea80.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "73f61e2cd00cde53ce12f752c610dbf080b9e10391caf0e24b42b98c89133ee1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "318beecdb68f8cf08c75c9db938b3aa7a4d1350ebdefa6d90c0c452ae1f715e0", "generation_seconds_batch": 21.88251304801088, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9508941173553467, "target_z": 0.18261715689108401, "ce": 6.09409761428833, "genuineness": 4.243753433227539, "identity": 0.6808252334594727, "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_CART_reg/73f61e2cd00cde53ce12f752c610dbf080b9e10391caf0e24b42b98c89133ee1.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "7408c2959d4a6a850d4f85a656fcf52ebf31e20968585806beb02ac7212ca3be", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2bdac85e752be079771333a7bfc88325d96118033b58094f30a78dd40a002a42", "generation_seconds_batch": 15.997995213023387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7409690618515015, "target_z": 0.02929688452277386, "ce": 6.308549880981445, "genuineness": 4.331617832183838, "identity": 0.45026177167892456, "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_CART_reg/7408c2959d4a6a850d4f85a656fcf52ebf31e20968585806beb02ac7212ca3be.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "74817074cfde151e2ebe4ffa7562244fa3ca6b792c2d7dd6798a085514a8f347", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9a84cf40390acc3a1326bbb31a14683b7c63c08000020c2be01a46a2d818f57d", "generation_seconds_batch": 15.506361396051943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0645480155944824, "target_z": 0.2656250979485311, "ce": 5.855817794799805, "genuineness": 4.143336772918701, "identity": 0.35078486800193787, "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_CART_reg/74817074cfde151e2ebe4ffa7562244fa3ca6b792c2d7dd6798a085514a8f347.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": "74c4f651c2d59acd2b53253b5039975efad9d0b4cb2dd465c7a0fc20179f450b", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e3ebb2da6a6a78eb93eae0a873a8301d4b56416a378ea884aaae0f871408259d", "generation_seconds_batch": 16.43756841495633, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8023229837417603, "target_z": -0.6562500217663403, "ce": 6.324435234069824, "genuineness": 4.582659721374512, "identity": 0.6090357303619385, "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_CART_reg/74c4f651c2d59acd2b53253b5039975efad9d0b4cb2dd465c7a0fc20179f450b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "751e433c49e4e9435cd0137a10c44e0ffd1039d9f916e2d1ffc04bdae6d93bc6", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "6e104e29adaf5c7b65e5892267d9699095d7b8870363c5803a10b71b1de0b545", "generation_seconds_batch": 19.627846773015335, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7075414657592773, "target_z": 0.004882799576235475, "ce": 6.292664527893066, "genuineness": 4.419482707977295, "identity": 0.2859436869621277, "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_CART_reg/751e433c49e4e9435cd0137a10c44e0ffd1039d9f916e2d1ffc04bdae6d93bc6.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": "752d78c8a5826d060b4d005f15b91c70808c47eaefd69e707274c8bf1b4b378f", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d822969c23136d66b07903886b3fe3349d72e651acd509bdaa014b20076d2567", "generation_seconds_batch": 20.212558539002202, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6097801923751831, "target_z": -0.7968750326495104, "ce": 6.268836498260498, "genuineness": 4.293961524963379, "identity": 0.5094723701477051, "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_CART_reg/752d78c8a5826d060b4d005f15b91c70808c47eaefd69e707274c8bf1b4b378f.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "75810b3a8f05e448f27511fd5cf2c507f5566bb41991d77156a26102d8924151", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "34bf302244d07e6fddbdb8f74da2e9121b644813e515d14c9d8f0d926cdb59c0", "generation_seconds_batch": 19.022186764981598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.235697031021118, "target_z": 0.39062501088317014, "ce": 5.903473377227783, "genuineness": 4.124508857727051, "identity": 0.7728230953216553, "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_CART_reg/75810b3a8f05e448f27511fd5cf2c507f5566bb41991d77156a26102d8924151.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "75907f53c6e37c47c3c282b327032ebaf7a2f9405dd7c7fe9f8c2c269865224f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "98ccff4dd07b059d406176e25e860e3b817bf0a71e991c53fdfb4893c4091d46", "generation_seconds_batch": 14.821375350933522, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4722113609313965, "target_z": -0.1669922330732749, "ce": 6.213237762451172, "genuineness": 4.457139015197754, "identity": 0.5546096563339233, "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_CART_reg/75907f53c6e37c47c3c282b327032ebaf7a2f9405dd7c7fe9f8c2c269865224f.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": "75981f0a718101ac7c6131c8996bf9e07d46e823ba7aa74d1d238997526dfd5d", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f7d72ce60de0bb0cb44a4749692d815da39e70cd5183bd039d1c472195361d97", "generation_seconds_batch": 14.940477830008604, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2863541841506958, "target_z": -0.30273435731484855, "ce": 6.141753673553467, "genuineness": 4.632867813110352, "identity": 0.708014965057373, "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_CART_reg/75981f0a718101ac7c6131c8996bf9e07d46e823ba7aa74d1d238997526dfd5d.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "75bb1de2f8a5c7715724772a41438850c91904b5cf8d7160bd816ee4db6237a1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a4d833b1680084a2e045296bc00b7f8332cd77d4eef91f84eed43536efab724c", "generation_seconds_batch": 21.406358398031443, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6100857257843018, "target_z": 0.6640624836752448, "ce": 5.847874641418457, "genuineness": 3.9425034523010254, "identity": 0.5792837738990784, "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_CART_reg/75bb1de2f8a5c7715724772a41438850c91904b5cf8d7160bd816ee4db6237a1.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": "75bd1e2e732dcc750573a96a7088fcac90e719387ba3df249329e96a070a8736", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "38830e80938f06ece7080c4f07920956c34a9ff227a684aa95574f6ec2351a06", "generation_seconds_batch": 15.685390901053324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.711400032043457, "target_z": -0.7226562527207925, "ce": 6.09409761428833, "genuineness": 4.482243061065674, "identity": 0.7786663770675659, "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_CART_reg/75bd1e2e732dcc750573a96a7088fcac90e719387ba3df249329e96a070a8736.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "75ee9dec2b4db4b478178a4fdcb7e45b34f6340defe7895800f7271398539c83", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "67a78c9330445304b10ce9cea94826f69077d822fbb91913c59a975f5106aef5", "generation_seconds_batch": 19.910247559077106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0618736743927, "target_z": 0.26367187363960376, "ce": 6.157639026641846, "genuineness": 3.9927117824554443, "identity": 0.480959415435791, "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_CART_reg/75ee9dec2b4db4b478178a4fdcb7e45b34f6340defe7895800f7271398539c83.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "efa7d71e660da8062dd4b9e7726629b614fc6230864e8c99ec2bf0aee9fd0a76", "generation_seconds_batch": 14.667770802974701, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.2781788110733032, "target_z": -1.039062483675245, "ce": 6.181467056274414, "genuineness": 4.6956281661987305, "identity": 0.6967223882675171, "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_CART_reg/77962898fb95ddc0e92a153772191fe01482e40f98d2245158cb92652e8a93c1.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "7798c8af4ef31ab73348e60ca8823a1e4591a33ba4e2ea8ce8c33b03a7163b6e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b427b398f4c10d77f41cc3f3b9f13e276b548853db5bffe872005bc83ec42898", "generation_seconds_batch": 15.528549426002428, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.674266815185547, "target_z": 0.7109376033901162, "ce": 5.982900142669678, "genuineness": 3.9927117824554443, "identity": 0.6808547973632812, "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_CART_reg/7798c8af4ef31ab73348e60ca8823a1e4591a33ba4e2ea8ce8c33b03a7163b6e.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "78fb90a3d7f6359a43479f43b5be69d94217e3603538a82cb9709547609cb993", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0dd60aefdc815aeb1fecb97dcbb40a5ad80a3b3609acb0caf7c7d07f8717cca8", "generation_seconds_batch": 15.20800564205274, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0418171882629395, "target_z": 0.2490234749108973, "ce": 6.3641486167907715, "genuineness": 4.030367851257324, "identity": 0.3798329830169678, "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_CART_reg/78fb90a3d7f6359a43479f43b5be69d94217e3603538a82cb9709547609cb993.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "4bd85fd89c398b1a3b3d11efba39b528489ea69fe0d7bca2a1d5d9567bc3abc1", "generation_seconds_batch": 15.212075505987741, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8558071255683899, "target_z": -0.6171874945584149, "ce": 6.1894097328186035, "genuineness": 4.381826400756836, "identity": 0.6855787038803101, "hypothesis": "De sleutels liggen op de keukentafel. Je keek in de la en daarna naast het raam. Neem ze mee als je weggaat.", "wer": 0.045454545454545456, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/794d3b9296e3c4be7c1e4182a2fa3cf095f7d593990763ce0f19c1f5e499a989.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "79b0705cc1e5f12be1ccda0bc349675818c6896b0c4678db0820a320f67ddf1e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9bb0976924321fd4641f38a1fdb66aef3747d2c29e626c4b612e4ac07a0b68f8", "generation_seconds_batch": 14.97264905704651, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0217604637145996, "target_z": 0.23437490205146885, "ce": 5.863759994506836, "genuineness": 3.923675537109375, "identity": 0.48670995235443115, "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_CART_reg/79b0705cc1e5f12be1ccda0bc349675818c6896b0c4678db0820a320f67ddf1e.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "79e9f49752dadb7dadb7d96e6d57557c47542543029e84a99c5ceed7e3c77c42", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "016833b2e428d359cf55a4c27ea6f4c7b993ba45fc0897a3b04aa3e3facefac2", "generation_seconds_batch": 19.71047442907002, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.856112480163574, "target_z": 0.8437498911682987, "ce": 5.272031307220459, "genuineness": 3.587907314300537, "identity": 0.21482731401920319, "hypothesis": "Ik heb een kleine noosje gevonden. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.1111111111111111, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/79e9f49752dadb7dadb7d96e6d57557c47542543029e84a99c5ceed7e3c77c42.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "7a48514cc30b51681006a8aef47b9b0647386999995dadcafc5637e37ac503e0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "55e684ac68c2d44ba1d996bb0ba05023cc454138946aca7687317c326271be1a", "generation_seconds_batch": 20.884523550979793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6528730392456055, "target_z": 0.6953125054415851, "ce": 5.657250881195068, "genuineness": 3.986435890197754, "identity": 0.533219575881958, "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_CART_reg/7a48514cc30b51681006a8aef47b9b0647386999995dadcafc5637e37ac503e0.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "7afc9884576a737a26206c7e9a6696662381b25d17198e2112fd46df89b1ccd8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "dbffc55863c6f2902ad6c95900b755789d3fab63eb4348cefa6d7d6fe84189bf", "generation_seconds_batch": 15.74608294700738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5245110988616943, "target_z": 0.6015624401425643, "ce": 5.903473377227783, "genuineness": 3.9173994064331055, "identity": 0.6109163761138916, "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_CART_reg/7afc9884576a737a26206c7e9a6696662381b25d17198e2112fd46df89b1ccd8.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "7b1596d89f00a8b68e12512c7b7cf7dbfb2fcbc050fb6814eb911c9356337b56", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "53e28bb2318d275975dcb7a8cbe8b8bf2212d81685cd08d4f66077b8fd517acc", "generation_seconds_batch": 16.022334815002978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.184887170791626, "target_z": 0.3535156644514917, "ce": 6.022613525390625, "genuineness": 4.231201171875, "identity": 0.5442192554473877, "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_CART_reg/7b1596d89f00a8b68e12512c7b7cf7dbfb2fcbc050fb6814eb911c9356337b56.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "7b6bb595058e7a309899ce3f375d195412318bf2dedcff94675b0c51b60747bb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7d0411fb3a7785b2f9656f8b0c86b79225063a4f3542c653071f9461809aefe2", "generation_seconds_batch": 14.970702097052708, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3159232139587402, "target_z": 0.44921877992871784, "ce": 5.855817794799805, "genuineness": 4.055472373962402, "identity": 0.28316283226013184, "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_CART_reg/7b6bb595058e7a309899ce3f375d195412318bf2dedcff94675b0c51b60747bb.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "7bcc6b2c5ea4acdc45c472d5b8070bd0a59a44973f4f0273a81591ecf670abcb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "eb7b064b0eb9dc9b3aee9095ead0ca71ca498c2e93e219c72f71f23d8e9bd217", "generation_seconds_batch": 19.634755764971487, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4120417833328247, "target_z": -0.2109375163247552, "ce": 6.09409761428833, "genuineness": 3.8420867919921875, "identity": 0.20765988528728485, "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_CART_reg/7bcc6b2c5ea4acdc45c472d5b8070bd0a59a44973f4f0273a81591ecf670abcb.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "7bf5e54b62757ffc01043234c70fe2fcfbf13c04816f2f2e386de7a3bdc95846", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "44c1437dc31849f5ea753141d81476dd22582fa9f5c7982b1dce84a655ca125f", "generation_seconds_batch": 20.382365124998614, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.663569927215576, "target_z": 0.7031250544158506, "ce": 5.446769714355469, "genuineness": 4.005263805389404, "identity": 0.5238097906112671, "hypothesis": "La puerta sigue abierta y he dejado tu abrigo sobre la silla. Dime cuándo estás listo.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/7bf5e54b62757ffc01043234c70fe2fcfbf13c04816f2f2e386de7a3bdc95846.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "7ccc714e27881ab9cad418c4ed0882cdc28006be2abeb2ae549c6caa28a4e86e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7fc2440c106d34a9dd87b593650ce92f3811a744fb8a8182ff44b9982f74db7b", "generation_seconds_batch": 19.803269280004315, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0162594318389893, "target_z": -0.5, "ce": 6.054384231567383, "genuineness": 4.406930446624756, "identity": 0.587341845035553, "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_CART_reg/7ccc714e27881ab9cad418c4ed0882cdc28006be2abeb2ae549c6caa28a4e86e.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "7ede02e88da349ae39748f440fcce29a573bc2b40ee7863eecd4162ffbc7e518", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "66e432beaaf63292accfa256e1c8c752d73748a8299329106fdc9c6dddacf442", "generation_seconds_batch": 18.998006792971864, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6607428789138794, "target_z": -0.02929688452277386, "ce": 6.157639026641846, "genuineness": 4.256305694580078, "identity": 0.039671365171670914, "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_CART_reg/7ede02e88da349ae39748f440fcce29a573bc2b40ee7863eecd4162ffbc7e518.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "7f0e5893584c06290db451891e7a96e8c7b703a8deefcb7977c1e14d477cedd0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2b92d0eced54b5f0c07c250f737533879970f955b342b5dadc99e9821a780281", "generation_seconds_batch": 15.884939100942574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.200932264328003, "target_z": 0.36523431378216803, "ce": 5.692992687225342, "genuineness": 4.432034492492676, "identity": 0.582207202911377, "hypothesis": "Trouver la petite boîte Elle était près de la vie, 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_CART_reg/7f0e5893584c06290db451891e7a96e8c7b703a8deefcb7977c1e14d477cedd0.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "8f54bd5b6210b6e02f949f67fc726bc8b7d5e7cb660e81c02db26b0062112f0f", "generation_seconds_batch": 15.879502214025706, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4386310577392578, "target_z": -0.9218750326495104, "ce": 6.38797664642334, "genuineness": 4.369274139404297, "identity": 0.3365117609500885, "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_CART_reg/7f6fce1a6f733719f45a363f28fee4e24b2f07a3396aa07b42392913d7c24239.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "7fae601a04b671d62abf083354dccd77dadbdd95106ce760e6f260875b790d10", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2e9db7164630279e2b067317521ec405988153b532a769deb8a195b98acf41e5", "generation_seconds_batch": 19.749617695924826, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1152050495147705, "target_z": -0.42773435731484855, "ce": 6.149696350097656, "genuineness": 4.413206577301025, "identity": 0.3591601252555847, "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_CART_reg/7fae601a04b671d62abf083354dccd77dadbdd95106ce760e6f260875b790d10.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "7fdc8de45f5c4810594d87cbfc35695ed8c32bc9fc6f3b93d197feacb471f1b1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8d23aac6bfdfe5387e5a40791335529620764af93be4dc10a525ea96c77dfd32", "generation_seconds_batch": 16.569732006057166, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6259782314300537, "target_z": -0.054687494558414936, "ce": 6.165581703186035, "genuineness": 4.143336772918701, "identity": 0.3761143088340759, "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_CART_reg/7fdc8de45f5c4810594d87cbfc35695ed8c32bc9fc6f3b93d197feacb471f1b1.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "19074c4baa93bdee8c72e962d10bd3ffed8261e3b7852a0587c18cf3b9fa2ebe", "generation_seconds_batch": 14.59807966905646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5135087966918945, "target_z": -0.8671875380910954, "ce": 6.340320587158203, "genuineness": 4.049196243286133, "identity": 0.6080166697502136, "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_CART_reg/80c2597a6b8f43b5338a6b37d654868d679f46ded48514b42837ab18fba327ad.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "80e967d1c61df08702f88ee4d549230c1f6b51eae9f66a756d0ae31ebaad8c21", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4e9f31b7de21c38bb6663a07d1955dbcc403f8eee0b2c4a99ea789d1cb17c08b", "generation_seconds_batch": 25.025240330956876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2838327884674072, "target_z": 0.42578130713664314, "ce": 5.911416053771973, "genuineness": 3.936227321624756, "identity": 0.3340783715248108, "hypothesis": "The door is still open. I put your code 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_CART_reg/80e967d1c61df08702f88ee4d549230c1f6b51eae9f66a756d0ae31ebaad8c21.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "8292fe4b916f873f41759a78859c8dd69b9d422c936753cc1f2d6ed6409f32e7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "34f9dd6cbfc944b1b32d91dcf4cc240812edcf44d551b547c26f7877b1cd77ba", "generation_seconds_batch": 20.303611007984728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6072587966918945, "target_z": -0.06835936819801867, "ce": 6.1894097328186035, "genuineness": 4.331617832183838, "identity": 0.35099560022354126, "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_CART_reg/8292fe4b916f873f41759a78859c8dd69b9d422c936753cc1f2d6ed6409f32e7.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "833c54fbd1960d23873a086423fc484cd2d333d921766bfa721c5f768d50807a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "45283101a82e070d2f697d671b2f6e286985e03fad4dc36b8d3ea12f627a15bf", "generation_seconds_batch": 21.9035704130074, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1982581615448, "target_z": 0.3632812636039627, "ce": 5.708878040313721, "genuineness": 3.9550557136535645, "identity": 0.529488742351532, "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_CART_reg/833c54fbd1960d23873a086423fc484cd2d333d921766bfa721c5f768d50807a.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": "83d10df0a53ad96c7aff88c91bb115fbca05e329061deba4b78cdb9faa061a25", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The keys are on the kitchen table.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "201765836567e5696dd90f714379a37720b692d26b2eb9af8183448ee444963e", "generation_seconds_batch": 15.7713026299607, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8504586815834045, "target_z": -0.6210937690455477, "ce": 6.403861999511719, "genuineness": 4.13078498840332, "identity": 0.359343022108078, "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_CART_reg/83d10df0a53ad96c7aff88c91bb115fbca05e329061deba4b78cdb9faa061a25.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "840b66a6a9fa7a9441727b1a8651e12fd3bf7e28a035476c4b0843131e141bf3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "74c847ae58854d1ae926cfa3a040ed48002bbe203bda435b715a066c5c3eaa55", "generation_seconds_batch": 15.74608294700738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.834718942642212, "target_z": 0.8281249673504896, "ce": 5.657250881195068, "genuineness": 3.6475296020507812, "identity": 0.12583765387535095, "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_CART_reg/840b66a6a9fa7a9441727b1a8651e12fd3bf7e28a035476c4b0843131e141bf3.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "840f9689be8caf40a4634f53c0635a273fd293df0e7cefa831e691b0ea6ece0b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7292e0ce3c2c5b87230e9130d6382a59aca2dc966672d9d039a7545719bb0ad9", "generation_seconds_batch": 19.60315748094581, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6353379487991333, "target_z": -0.04785155773861307, "ce": 5.847874641418457, "genuineness": 4.2751336097717285, "identity": 0.5876897573471069, "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_CART_reg/840f9689be8caf40a4634f53c0635a273fd293df0e7cefa831e691b0ea6ece0b.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "84217be9e7231792ded6bdf431ee2a4dfddbd2297a46b359ac1c45dec372fcc4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5792b98a8233ad44ad2393944f826ee2ae870ea731545d94c6b1e647c8d19445", "generation_seconds_batch": 22.19703052204568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1660149097442627, "target_z": -0.39062501088317014, "ce": 6.125868320465088, "genuineness": 4.268857479095459, "identity": 0.7152964472770691, "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_CART_reg/84217be9e7231792ded6bdf431ee2a4dfddbd2297a46b359ac1c45dec372fcc4.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "844ed0d0ab5535518a4c945814007fa72893b0896278eb6c6891bf874bd4dfb1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "583aa6de5fd48d82f3d4c2b23991433e933e354aef86ebb48819985f19c69d1a", "generation_seconds_batch": 16.429197776014917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3239457607269287, "target_z": 0.455078104594056, "ce": 6.078212261199951, "genuineness": 4.18099308013916, "identity": 0.2796230912208557, "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_CART_reg/844ed0d0ab5535518a4c945814007fa72893b0896278eb6c6891bf874bd4dfb1.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "849e0cd1e5e4be2d61f7290a42a626252c6388937e3e3f738878b43f69881d16", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c1337f4c9674e62cc205c4ce0c02780b29e484fec0a0f665cdd80efb209231c5", "generation_seconds_batch": 21.276315052993596, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2838327884674072, "target_z": 0.42578130713664314, "ce": 6.24500846862793, "genuineness": 4.300237655639648, "identity": 0.3458055257797241, "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_CART_reg/849e0cd1e5e4be2d61f7290a42a626252c6388937e3e3f738878b43f69881d16.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "84a58550375640ed98a4c19f0ac141d098e1fee680fa44807af2f38ff923ac7d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "de20f2a87b8a84882e466c1d3560c703954c5cace38242679e7564ef73921c4f", "generation_seconds_batch": 19.540031381999142, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.551253318786621, "target_z": 0.6210938125782283, "ce": 5.613565921783447, "genuineness": 4.20609712600708, "identity": 0.4364045262336731, "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_CART_reg/84a58550375640ed98a4c19f0ac141d098e1fee680fa44807af2f38ff923ac7d.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "850521e1bfbe85191e0db2df89db053d05e426e7609c106d4dd02de20920dd51", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "77561825f17fd16238a7629dc50059564201aec34ce570a6734f02961dcb1a49", "generation_seconds_batch": 20.011779442080297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1822128295898438, "target_z": 0.3515624401425643, "ce": 5.998785495758057, "genuineness": 4.231201171875, "identity": 0.21011240780353546, "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_CART_reg/850521e1bfbe85191e0db2df89db053d05e426e7609c106d4dd02de20920dd51.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "854f68b6a55628f00e2830de3149ecc4f0965ab22eeaa185b3a4b8602de529bc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7888ae1c7b05481b46bc93b7fa674082d29722276a8d6fb6dba199d715e7a734", "generation_seconds_batch": 22.259696130058728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8614611625671387, "target_z": 0.8476563397861535, "ce": 5.7406487464904785, "genuineness": 3.7667744159698486, "identity": 0.07528695464134216, "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_CART_reg/854f68b6a55628f00e2830de3149ecc4f0965ab22eeaa185b3a4b8602de529bc.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": "85893615ef65b119c88e96fd27a31816c21ecd4fea6c83d2fb9c578a96b62637", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "afa8d0594f62bb1bc21004e97b0264c1b1c9e38dd6182272a7a846971e6d26d7", "generation_seconds_batch": 20.631749409018084, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.74899160861969, "target_z": 0.03515620918811203, "ce": 6.260893821716309, "genuineness": 4.162164688110352, "identity": 0.5319669246673584, "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_CART_reg/85893615ef65b119c88e96fd27a31816c21ecd4fea6c83d2fb9c578a96b62637.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "8624c7f90eebe4c76dc45c1ee91415dd05641170eb959cda323c9a05602fb1ac", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "859af23124afc18ed0357f5b7d97ce4a4b6b16ee88b6141882f839347e6db787", "generation_seconds_batch": 20.83032883494161, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.925336480140686, "target_z": -0.5664062309544523, "ce": 6.078212261199951, "genuineness": 4.055472373962402, "identity": 0.4849799573421478, "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_CART_reg/8624c7f90eebe4c76dc45c1ee91415dd05641170eb959cda323c9a05602fb1ac.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "863cf9b477ac8c43f5d6101c8ebeebb2aac81f4676afc8768532bb164d9bef9c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f085bfd5db003ac1a6075b46d2c26a28560cccab4ad0ba63b18e6784170cbe15", "generation_seconds_batch": 19.98434087098576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9841689467430115, "target_z": -0.5234375163247552, "ce": 6.229123115539551, "genuineness": 4.319066047668457, "identity": 0.5889873504638672, "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_CART_reg/863cf9b477ac8c43f5d6101c8ebeebb2aac81f4676afc8768532bb164d9bef9c.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "86c8e688b7c16c472e346480f81f73db5b7fe2e02774a17901c32bf4a624ce81", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c39bdda0568556a349bb9d5edc1369b340445b4d243a0ce98a1ff581e072addb", "generation_seconds_batch": 19.759780065971427, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7488389015197754, "target_z": -0.6953125054415851, "ce": 6.1020402908325195, "genuineness": 4.256305694580078, "identity": 0.660589873790741, "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_CART_reg/86c8e688b7c16c472e346480f81f73db5b7fe2e02774a17901c32bf4a624ce81.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "8714fae6fcacafa218cc3b3d11e26ac49759ea296379c754fb78c085071f5ecc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "504e99bb1207063292663829e1c4e78650908c35952825974d2f0278aff2b00b", "generation_seconds_batch": 19.168484138092026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.09128999710083, "target_z": 0.28515629625347305, "ce": 6.0305562019348145, "genuineness": 4.15588903427124, "identity": 0.22920985519886017, "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_CART_reg/8714fae6fcacafa218cc3b3d11e26ac49759ea296379c754fb78c085071f5ecc.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "87d912da9dcb11d3673df07d6021c93978c104e2c1e1e1d3ea6b8a8fca6e6c0d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6ba8a2ab3ff7eecb87b0aa75e2bc7f395a7e8d4be3c40016d85c8784159d36a2", "generation_seconds_batch": 19.865049802931026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7583513259887695, "target_z": 0.04199214600791389, "ce": 6.133810997009277, "genuineness": 4.469690799713135, "identity": 0.49000298976898193, "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_CART_reg/87d912da9dcb11d3673df07d6021c93978c104e2c1e1e1d3ea6b8a8fca6e6c0d.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "88001a37cf3e436fecd4c81de4a8d2dba3b0442d34933406a560eb448fcc2a33", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "949d8f629a6a43bfb05ab67ca91209ffae291c7a2ede6cd6454b1ee9fdbe1372", "generation_seconds_batch": 18.76072395790834, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.377429962158203, "target_z": 0.4941406753346618, "ce": 5.657250881195068, "genuineness": 4.005263805389404, "identity": 0.45540517568588257, "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_CART_reg/88001a37cf3e436fecd4c81de4a8d2dba3b0442d34933406a560eb448fcc2a33.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": "8815403757ce76358e12b6ac9589ab14a9b96db4f7486339197bb7dd68c78469", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "d7fc613e8c399536a366071801a21be2b572e6e4399df8d6ee55278dde2baaad", "generation_seconds_batch": 20.631749409018084, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5830382108688354, "target_z": -0.8164062309544523, "ce": 6.2767791748046875, "genuineness": 4.381826400756836, "identity": 0.45229846239089966, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/8815403757ce76358e12b6ac9589ab14a9b96db4f7486339197bb7dd68c78469.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "8825fa6be260536b81c5746366faa6361780eaf9794caaa6df54ff32ef094777", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3b6c5468d6a0e8386bf9cebc61caba2d9c3e5b62c212b254c942c9b55df633e3", "generation_seconds_batch": 15.883290670928545, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1822128295898438, "target_z": 0.3515624401425643, "ce": 5.911416053771973, "genuineness": 3.898571252822876, "identity": 0.45341384410858154, "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_CART_reg/8825fa6be260536b81c5746366faa6361780eaf9794caaa6df54ff32ef094777.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "88768bc6f3c1b352e4fb6a7336477752015100a756b23e39483e81fa52d7b4a9", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "898c1d578e7b5aec226ae83639dd35eaad3248e63d53c16727c765e5f4d918b8", "generation_seconds_batch": 16.510753549053334, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8265435695648193, "target_z": 0.09179684099009336, "ce": 6.125868320465088, "genuineness": 4.137060642242432, "identity": 0.49912625551223755, "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_CART_reg/88768bc6f3c1b352e4fb6a7336477752015100a756b23e39483e81fa52d7b4a9.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "88772e62b487d8ba742028795b8061c58f7b0cedc9cb25de8ee5cc8ee7c6d61f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "46e8e5743c28bc7a5b9484da620505e64f1a86e99221f0d9e9883b6631fc4462", "generation_seconds_batch": 16.36496934306342, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6206297874450684, "target_z": -0.05859376904554772, "ce": 6.324435234069824, "genuineness": 4.256305694580078, "identity": 0.40648961067199707, "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_CART_reg/88772e62b487d8ba742028795b8061c58f7b0cedc9cb25de8ee5cc8ee7c6d61f.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "890c3e47d9a69ae2697f7e9935f8e39ee268f70b0dcdc25dd4309cfd89659b01", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9bf38587eaa0c11ce8a0a9482ad2251543cc61eb39a8a6b9675098bb8e7f09ea", "generation_seconds_batch": 24.037589969928376, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0030412673950195, "target_z": 0.22070320254258716, "ce": 5.764476776123047, "genuineness": 4.068024158477783, "identity": 0.5280943512916565, "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_CART_reg/890c3e47d9a69ae2697f7e9935f8e39ee268f70b0dcdc25dd4309cfd89659b01.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "8957e6a942462ade80ff7dbe850f402c7c2216578f19c3fa390ce2b853552ff0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "18cabb95f41dec4ecb03f01b34afa9c849386bca1f2087287c9e56d22059871b", "generation_seconds_batch": 19.43283155199606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8733421564102173, "target_z": 0.1259765250891027, "ce": 6.0305562019348145, "genuineness": 4.444586753845215, "identity": 0.4423895478248596, "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_CART_reg/8957e6a942462ade80ff7dbe850f402c7c2216578f19c3fa390ce2b853552ff0.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": "89591af3279e65e764e27c5daac280b358266915b8b79c1e49632a979274a907", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2a2ccb14e0099f199a93ee951c9aa09cb38b241ab6a21e85db6ceff6241dcfea", "generation_seconds_batch": 15.465044792043045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9520785212516785, "target_z": -0.5468749891168299, "ce": 6.260893821716309, "genuineness": 4.3881025314331055, "identity": 0.6944324970245361, "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_CART_reg/89591af3279e65e764e27c5daac280b358266915b8b79c1e49632a979274a907.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "8ab92625458bd3fc0aa7647acb3fdf47529d56b68fb667ff50e331285aa149e0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a3726c110150f1711899d0b099db3d8e0335675326691b5da01124fbd6f78cbd", "generation_seconds_batch": 19.66804720496293, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.856112480163574, "target_z": 0.8437498911682987, "ce": 4.627682209014893, "genuineness": 3.402764081954956, "identity": 0.12479820847511292, "hypothesis": "Ik heb het kleine loonje gevonden. Het stond naast het oude graam. We kunnen er samen in kijken.", "wer": 0.1111111111111111, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/8ab92625458bd3fc0aa7647acb3fdf47529d56b68fb667ff50e331285aa149e0.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "8b123400f57df0931e390e51986636ed8e2a6c2b52dd3f213e9380d304ce29a4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "bd69cf09b2675039d67db0a3e54e7b567bde7b14171c98d60d188e1eb041c415", "generation_seconds_batch": 21.806784193962812, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3801040649414062, "target_z": 0.49609372551286723, "ce": 5.974957466125488, "genuineness": 3.8044304847717285, "identity": 0.4206894636154175, "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_CART_reg/8b123400f57df0931e390e51986636ed8e2a6c2b52dd3f213e9380d304ce29a4.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "8c9b3e125eb3ff7b5060b0ca0db5528a4219127b52159e5cdee9314633045d7c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "33f161abac00a4d89bac8da3f92c3e5f52485d2ca702ba739824b1a805487fed", "generation_seconds_batch": 19.635649631964043, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2890284061431885, "target_z": -0.30078122007128216, "ce": 6.284721851348877, "genuineness": 4.3504462242126465, "identity": 0.41889193654060364, "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_CART_reg/8c9b3e125eb3ff7b5060b0ca0db5528a4219127b52159e5cdee9314633045d7c.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "8d1ad42f8e3668e5af7bceca3d5b0374f51b9ddb7988095debe2fe3494a1ea16", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2467eaa4d73aec2aa6ad6a0741c637c50e87822c1e660fbe9bf0e14dd957e19e", "generation_seconds_batch": 23.27186566707678, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1981053352355957, "target_z": -0.36718753809109544, "ce": 5.788304805755615, "genuineness": 4.3127899169921875, "identity": 0.44729915261268616, "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_CART_reg/8d1ad42f8e3668e5af7bceca3d5b0374f51b9ddb7988095debe2fe3494a1ea16.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "8e1ef9d237a8114ec6bd8b4fe5087d7a3ae16ce02e07ea045da31c37df52fb79", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "344cb3b55048e87fa8c40a3b3da9206f2ac285ea7a537723b1ac96411af03733", "generation_seconds_batch": 19.882928965962492, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2544164657592773, "target_z": 0.40429688452277385, "ce": 6.070269584655762, "genuineness": 3.8922953605651855, "identity": 0.497437059879303, "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_CART_reg/8e1ef9d237a8114ec6bd8b4fe5087d7a3ae16ce02e07ea045da31c37df52fb79.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "8e35d0265ac42351869b7e35e481e369b08b2607df680c771143259519c61ec5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2d4a2bc30101579fa6ec4e080df490e9a225620e607a5494836e030cd38ae69b", "generation_seconds_batch": 16.355668568983674, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2676347494125366, "target_z": -0.3164062309544523, "ce": 6.213237762451172, "genuineness": 4.149612903594971, "identity": 0.3241012394428253, "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_CART_reg/8e35d0265ac42351869b7e35e481e369b08b2607df680c771143259519c61ec5.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "8e78a5891a5998f02e199e4c2eb57f508584cffa4924edd701fbb50d95d19918", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "664c4fece5a1a945a36e81175f9324e0d05e400926f9892e0c8034c33314bbfd", "generation_seconds_batch": 20.04269958799705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0323046445846558, "target_z": -0.4882812636039627, "ce": 6.292664527893066, "genuineness": 4.362998008728027, "identity": 0.5740088224411011, "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_CART_reg/8e78a5891a5998f02e199e4c2eb57f508584cffa4924edd701fbb50d95d19918.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "8e807d25bb950bf94df3ab73eb25231bf0999912226608d2f9cd7b4aa902c045", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0e4fb8df72ce1c3b275d6c24c7fb1527cec86d0c570032580e78a687836f0aca", "generation_seconds_batch": 19.940332753001712, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.088615655899048, "target_z": 0.28320307194454564, "ce": 6.0305562019348145, "genuineness": 4.413206577301025, "identity": 0.5527327060699463, "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_CART_reg/8e807d25bb950bf94df3ab73eb25231bf0999912226608d2f9cd7b4aa902c045.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "8ee3863519f593e373c6b0fdec07093eef93089ab95e5829d0fa7493c68a4294", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "43f82bd8b3c34dc2e9f6a1e486897d5607f336508f8d303c9290e003203271ed", "generation_seconds_batch": 19.540031381999142, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.770538091659546, "target_z": 0.7812500217663403, "ce": 5.254160404205322, "genuineness": 3.8546390533447266, "identity": 0.10928844660520554, "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_CART_reg/8ee3863519f593e373c6b0fdec07093eef93089ab95e5829d0fa7493c68a4294.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "8fca30f528577a0b521a0c267bef88b264ce3482c59e82f3299b1fa79a179e53", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f22f8a5706eb94b6a5a6afc4f479bc8d6cfe32e8b04f471b5bba3b21bad965db", "generation_seconds_batch": 15.105503295082599, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5138144493103027, "target_z": 0.5937500652990207, "ce": 5.6413655281066895, "genuineness": 3.7793264389038086, "identity": 0.6218491196632385, "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_CART_reg/8fca30f528577a0b521a0c267bef88b264ce3482c59e82f3299b1fa79a179e53.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "9008f7ff89c0b5f932e7629125819f6490a8a4f3d4456d00eac1b1ef32d9a70f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6ca543adc61e6f70248b7aa08c312c99eb75215eea1875bd241cb97fe9e28394", "generation_seconds_batch": 21.09462536103092, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9841689467430115, "target_z": -0.5234375163247552, "ce": 6.1894097328186035, "genuineness": 4.570107460021973, "identity": 0.6755891442298889, "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_CART_reg/9008f7ff89c0b5f932e7629125819f6490a8a4f3d4456d00eac1b1ef32d9a70f.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": "9069bf3a646413fe8d9a87cbecf0abc09aeef8c926d2e6f0ab3c523bb38c8765", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Escribí el número en el sobre.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ceffe759283ea3c8a7cdd61cf6ada6d8867e2b111dfe74a646ba412ea39d2569", "generation_seconds_batch": 20.051023079897277, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0938113927841187, "target_z": -0.4433593681980187, "ce": 6.09409761428833, "genuineness": 3.9927117824554443, "identity": 0.529815673828125, "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_CART_reg/9069bf3a646413fe8d9a87cbecf0abc09aeef8c926d2e6f0ab3c523bb38c8765.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "90ae094f253b10b9e49411df81b233c9d825b22ea1e81fc3fad05414ff09f7cb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9815a378b659409e69f9c69147d37a04041e77972dd7f5f4a4e6f2fe9eb70302", "generation_seconds_batch": 15.293078011018224, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1287288665771484, "target_z": 0.3125000435326805, "ce": 5.831989288330078, "genuineness": 3.986435890197754, "identity": 0.6509982347488403, "hypothesis": "De deur staan nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "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_CART_reg/90ae094f253b10b9e49411df81b233c9d825b22ea1e81fc3fad05414ff09f7cb.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "90d92729196b325e94d51ed78379237571760398185336e3326169e3fbf3aa72", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "6175a27a5a8c6b347934c409735ef81b4337b3680d606776b2ae84e1f3d366cb", "generation_seconds_batch": 15.880435994942673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.246393918991089, "target_z": 0.3984375598574357, "ce": 5.982900142669678, "genuineness": 3.8797430992126465, "identity": 0.278265118598938, "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_CART_reg/90d92729196b325e94d51ed78379237571760398185336e3326169e3fbf3aa72.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "90fc5b9425b328bf47547a1aefc02b78fbb22365a64f505fa9a59c3f0984203d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "79477aa7ca26c953bf5a92df488b8ffc18db4084f9e75aa959035c83075aea22", "generation_seconds_batch": 20.504617818049155, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4335882663726807, "target_z": 0.535156296253473, "ce": 5.7247633934021, "genuineness": 4.281409740447998, "identity": 0.03789385035634041, "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_CART_reg/90fc5b9425b328bf47547a1aefc02b78fbb22365a64f505fa9a59c3f0984203d.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "914bbc6a81136154df8d9dca281043e440ded2f14a4c4900f04f9fff6f05763b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f1bdf3c7e1d804b15b85cee2b1ebb20b66d6b1c76268f3be86c747b5fbea3169", "generation_seconds_batch": 21.452150929020718, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.535207986831665, "target_z": 0.6093749891168299, "ce": 5.597680568695068, "genuineness": 3.798154592514038, "identity": 0.3634295165538788, "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_CART_reg/914bbc6a81136154df8d9dca281043e440ded2f14a4c4900f04f9fff6f05763b.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "917853789526fdf61864363129ee8bdf72a4463d543b7e8fb5f15d6c04300a19", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "59ab44ab091b09736c406e3826d26bfca926744c193d12da6da7d561977a406f", "generation_seconds_batch": 15.676996757974848, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6112700700759888, "target_z": -0.06542970586534959, "ce": 5.935244083404541, "genuineness": 4.055472373962402, "identity": 0.422946572303772, "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_CART_reg/917853789526fdf61864363129ee8bdf72a4463d543b7e8fb5f15d6c04300a19.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": "91ad0717a853d30f48b39cf255f680f926fdd2db49f464a3f9418a32d6fea18f", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a681b9f3d6b2a8940f1e168b7409c7619f576e74fa472d467945856e8584c664", "generation_seconds_batch": 15.426801213994622, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0563725233078003, "target_z": -0.47070311547722615, "ce": 6.038498878479004, "genuineness": 4.231201171875, "identity": 0.6369622349739075, "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_CART_reg/91ad0717a853d30f48b39cf255f680f926fdd2db49f464a3f9418a32d6fea18f.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": "91b64560a04d96f9164048d0132a94a6772d8c5ec9ca8f17100590514ccd2fb4", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Your message arrived after I had left.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "406081f9f46d74170b1e3a6de38ef43dc6b7211e7b45f8a39c2dc3fc1104c182", "generation_seconds_batch": 20.354535614023916, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9948657751083374, "target_z": -0.5156250108831701, "ce": 6.356205940246582, "genuineness": 4.193545341491699, "identity": 0.4235094487667084, "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_CART_reg/91b64560a04d96f9164048d0132a94a6772d8c5ec9ca8f17100590514ccd2fb4.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "91e9a9929419b888aa08eaedccfe16bb44a77cc633019f56e43916c73b6d4fe4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f9f019492814368a2cf74338a223494eb6f71f3b8754458b2739439cbe934803", "generation_seconds_batch": 21.779950855998322, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6099330186843872, "target_z": -0.06640623095445228, "ce": 6.1894097328186035, "genuineness": 4.319066047668457, "identity": 0.48901253938674927, "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_CART_reg/91e9a9929419b888aa08eaedccfe16bb44a77cc633019f56e43916c73b6d4fe4.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "9208b29578fe22d9e309ba67fab8268f956649b07e7cdbaed73d23a1d40eb0e6", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "ed69fe1d49520876e11399f3674371ea17fd6e85b12e7e7269afaa6ef5251410", "generation_seconds_batch": 24.18821112893056, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8293704986572266, "target_z": 0.8242186928633568, "ce": 5.246217727661133, "genuineness": 4.118232727050781, "identity": 0.5087409615516663, "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_CART_reg/9208b29578fe22d9e309ba67fab8268f956649b07e7cdbaed73d23a1d40eb0e6.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "92177731a0ac343d45a1b4414a4859a7b2543decd523d2472bfcdbeac5e09fdd", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ad381817e7b41a8bb1d455fd489ce80b9ee3a78c80bce280d42dc94a289039dd", "generation_seconds_batch": 19.71243744494859, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4014978408813477, "target_z": 0.5117188234613983, "ce": 6.133810997009277, "genuineness": 4.243753433227539, "identity": 0.3349494934082031, "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_CART_reg/92177731a0ac343d45a1b4414a4859a7b2543decd523d2472bfcdbeac5e09fdd.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "926920c76ddb2af5e951e33f07fcc8bd42df944da9b77327ef7d0d2e7136dbf7", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a461bf39f1d2e77b5d80c0ec691fadf1a014000595c501f83fcd3aef0a93d241", "generation_seconds_batch": 16.10707020701375, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1098566055297852, "target_z": -0.4316406318019813, "ce": 6.284721851348877, "genuineness": 4.356721878051758, "identity": 0.35441941022872925, "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_CART_reg/926920c76ddb2af5e951e33f07fcc8bd42df944da9b77327ef7d0d2e7136dbf7.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "92ba38cff0be3333de00c9f4fdff9538de09103a99d56c5bf09d7454559e4cca", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "58ce1f7ba0e22602ce1d0b4dab6d8123d9d51f639113d81142df49334d74ab28", "generation_seconds_batch": 19.27111886208877, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.956242561340332, "target_z": 0.1865234313782168, "ce": 5.816103935241699, "genuineness": 4.005263805389404, "identity": 0.27080193161964417, "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_CART_reg/92ba38cff0be3333de00c9f4fdff9538de09103a99d56c5bf09d7454559e4cca.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "930c37ae47e5abff70de4d0cf8623f9db92cc76072136af1da558e344e0e1064", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "86b5656c49d0744e8fcfb9dfeaebd17ee7792d5d6bf703ad5855dc14ca36175b", "generation_seconds_batch": 20.256401565042324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9615910053253174, "target_z": 0.19042970586534957, "ce": 6.23706579208374, "genuineness": 4.362998008728027, "identity": 0.6586346626281738, "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_CART_reg/930c37ae47e5abff70de4d0cf8623f9db92cc76072136af1da558e344e0e1064.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "93716ae2a19b87d4674be3d257e7fe987060c1b1d114976171e4e0a7f5b23ac9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "87147fcb88727699181c980f32741b15b1fbddc8c0b3d86198fb97129cc3c1e7", "generation_seconds_batch": 19.850642392993905, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0804402828216553, "target_z": -0.45312505441585066, "ce": 6.1894097328186035, "genuineness": 4.519899368286133, "identity": 0.7854541540145874, "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_CART_reg/93716ae2a19b87d4674be3d257e7fe987060c1b1d114976171e4e0a7f5b23ac9.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": "9384111b6cbe4b9796d50cfc938a27f8c0441a7d4689030921183ad0ab3e1aae", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The road turns left behind the bridge.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "94e3e2669e0d375e07b18d73293fcfb32f14e2f002fd8d90d3533ff7ea924ec8", "generation_seconds_batch": 21.164766506990418, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8130198121070862, "target_z": -0.6484375163247552, "ce": 6.403861999511719, "genuineness": 4.18099308013916, "identity": 0.4205578565597534, "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_CART_reg/9384111b6cbe4b9796d50cfc938a27f8c0441a7d4689030921183ad0ab3e1aae.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "94238c13f48a90b24af9a42c273e1d9854c5b5c6750268a8949848ca875dcf93", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "92337604a5a20f00ccc6e8ee6c6fea6372d4eeef730126359f5945692b0eb6e8", "generation_seconds_batch": 21.623023084015585, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8665039539337158, "target_z": -0.6093749891168299, "ce": 5.895530700683594, "genuineness": 4.256305694580078, "identity": 0.42106562852859497, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raamon. 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_CART_reg/94238c13f48a90b24af9a42c273e1d9854c5b5c6750268a8949848ca875dcf93.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "94760efc4df0d0236e00335ec7c92c463de0a6b346d7bade89f688592d4116f2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a6c91355d226d6f8e4fa8995842c2147e9babf73c0b3928b99741b7edfa10567", "generation_seconds_batch": 21.15716633806005, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.567298412322998, "target_z": 0.6328124619089046, "ce": 5.871703147888184, "genuineness": 4.137060642242432, "identity": 0.36104080080986023, "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_CART_reg/94760efc4df0d0236e00335ec7c92c463de0a6b346d7bade89f688592d4116f2.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": "956e599ed6d457484235fee1f23767a871d18b98a8c20793aeb2bff9082c5148", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "054b14f2d7739efe7fedb550d4ed49801b410a4438ea40c6fb6a8da1a912fb88", "generation_seconds_batch": 15.768150239018723, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8198580741882324, "target_z": 0.08691404141385788, "ce": 5.919358730316162, "genuineness": 4.118232727050781, "identity": 0.5529240369796753, "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_CART_reg/956e599ed6d457484235fee1f23767a871d18b98a8c20793aeb2bff9082c5148.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "956f9bd3bda32ecf2f22d9698c7ec3e4372357e27ac96df047c480967377fe92", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e0f1d31e4e82b41da0ca3accb1ed1a564e55b8e4db5a0a0d84389e66e3feb024", "generation_seconds_batch": 19.523576807929203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3961493968963623, "target_z": 0.5078125489742655, "ce": 6.006728172302246, "genuineness": 4.074300289154053, "identity": 0.15511156618595123, "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_CART_reg/956f9bd3bda32ecf2f22d9698c7ec3e4372357e27ac96df047c480967377fe92.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "95a846bfd397c9425d953b4a771761ceb436709ef67079c751f6f721b4927edf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a6178ca66d0188e61fffcd80968017b0bf474394dbe5bfb09eaefb131164baba", "generation_seconds_batch": 19.827381295035593, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0807456970214844, "target_z": 1.0078123748435435, "ce": 5.645336627960205, "genuineness": 3.9299514293670654, "identity": 0.24759453535079956, "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_CART_reg/95a846bfd397c9425d953b4a771761ceb436709ef67079c751f6f721b4927edf.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "960897569ce9b342a6f2a8631a5e8c2c4320f68fcc99154015db68516f7aae4d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f544d8271b0e02f4719b1bac12a7bb5ead1a69ab9f6cabca63711115e1797d1b", "generation_seconds_batch": 20.58554262900725, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8718523383140564, "target_z": -0.6054687581623776, "ce": 6.252951145172119, "genuineness": 4.482243061065674, "identity": 0.5792284607887268, "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_CART_reg/960897569ce9b342a6f2a8631a5e8c2c4320f68fcc99154015db68516f7aae4d.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "965e3f7cc10418df1d7d1371390b6f012f29c1623b3bd7d83320ba9234c187c1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "064d4a6c2d04250ac4203d2c219415556d3ebe5ab64d20937d0e4a22a440b693", "generation_seconds_batch": 20.031144544947892, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8665039539337158, "target_z": -0.6093749891168299, "ce": 6.23706579208374, "genuineness": 4.6705241203308105, "identity": 0.5766100883483887, "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_CART_reg/965e3f7cc10418df1d7d1371390b6f012f29c1623b3bd7d83320ba9234c187c1.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "96faf22fe0437ead0996e99451d478a859600db069f1aa609b436d1bf3a92c51", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "75b5f7c16b098a212f7d742deb6d3c3404d9ba127807d9d1ceb65cd06b60c443", "generation_seconds_batch": 21.301889894064516, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.059352397918701, "target_z": 0.9921876251564564, "ce": 5.935244083404541, "genuineness": 3.9173994064331055, "identity": 0.29327985644340515, "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_CART_reg/96faf22fe0437ead0996e99451d478a859600db069f1aa609b436d1bf3a92c51.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "971795638a49967a5112fe77d50c80bfa6ecbf0ff5068883578979fe37739647", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2a4fa6e1de104d80853f03a01a9ab6e707349f48ffa19254d805cd708dfb5d7f", "generation_seconds_batch": 22.259696130058728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.850764274597168, "target_z": 0.839843790811888, "ce": 4.96226692199707, "genuineness": 3.911123275756836, "identity": 0.1676105409860611, "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_CART_reg/971795638a49967a5112fe77d50c80bfa6ecbf0ff5068883578979fe37739647.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": "9778c03bbcb55dddf93dcf514ce870bef44683c8be58ec1ea52d52263984f831", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Escribí el número en el sobre.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3cd1a5d016f38e38d185c6dc399798c941c9a2413bdc0d14a2ceeff3195c7b9e", "generation_seconds_batch": 15.572665825951844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4508178234100342, "target_z": -0.18261715689108401, "ce": 6.062326908111572, "genuineness": 4.18099308013916, "identity": 0.4806099832057953, "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_CART_reg/9778c03bbcb55dddf93dcf514ce870bef44683c8be58ec1ea52d52263984f831.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": "98a65ff023c5a7552fe12b700a08fe6c4e4daa58e4e1c17d1810d55152a996de", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f22084d450f7e097281d8f0e3bf771ecfbd91a265a4023a760a1a90a8d298b41", "generation_seconds_batch": 20.814140757080168, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1472954750061035, "target_z": -0.40429688452277385, "ce": 6.0305562019348145, "genuineness": 4.595211505889893, "identity": 0.6959881782531738, "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_CART_reg/98a65ff023c5a7552fe12b700a08fe6c4e4daa58e4e1c17d1810d55152a996de.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "98e2883f3746cd2f7845d5977e50ffe7ab55b399eca6cb1cd2ab70e68913c259", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e804e467c82976e2eaa0d1126e36734c9da4048f00c4561c0cd325dad14f5865", "generation_seconds_batch": 19.41208502301015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.107335090637207, "target_z": 0.29687494558414934, "ce": 5.704906940460205, "genuineness": 3.848362922668457, "identity": 0.37758785486221313, "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_CART_reg/98e2883f3746cd2f7845d5977e50ffe7ab55b399eca6cb1cd2ab70e68913c259.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "99007d0239fcd48cb8a8f5db8f0e08ce66f27d74e27f5e13a64177b63f607fd9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "10e6089fe85c85cdd1b0ae6d1ee33b578cb912f9fa15f5b8ab9b0280db02b4be", "generation_seconds_batch": 20.83032883494161, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7636997699737549, "target_z": 0.04589842049504668, "ce": 5.911416053771973, "genuineness": 3.998987913131714, "identity": 0.5165372490882874, "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_CART_reg/99007d0239fcd48cb8a8f5db8f0e08ce66f27d74e27f5e13a64177b63f607fd9.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "9905a2f3e4a475b846c58532b0572420e380b867d74c75c20d1d55618d1b1ef4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "098946a8a5629deafe08466bcc2c51e21a9bba079bf51aa55efaa7457d13d137", "generation_seconds_batch": 20.89842421002686, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1795387268066406, "target_z": 0.3496093899643589, "ce": 6.173524379730225, "genuineness": 3.9927117824554443, "identity": 0.7515114545822144, "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_CART_reg/9905a2f3e4a475b846c58532b0572420e380b867d74c75c20d1d55618d1b1ef4.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "49b8d1b76705072c8d9234055451c3e237208da3e438c875893cd5adf5743498", "generation_seconds_batch": 15.435017101000994, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.47072160243988037, "target_z": -0.8984374727920746, "ce": 6.372091293334961, "genuineness": 4.65797233581543, "identity": 0.5392839908599854, "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_CART_reg/99bf1b7d3eed7b885fdae888ab00b4960791c3971d259b46f771306d2801e39b.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "99c6df11125801c5b62f5e20f8727afedfbb1a227911c6cf2153105f33a9ac75", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b03157c291f9ac9bc5aa6ef84bf79c0101dd3d2f8a851d2646c2885f2cbbe31d", "generation_seconds_batch": 18.996058237971738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6179555654525757, "target_z": -0.06054690628911411, "ce": 6.316492557525635, "genuineness": 4.413206577301025, "identity": 0.6988677978515625, "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_CART_reg/99c6df11125801c5b62f5e20f8727afedfbb1a227911c6cf2153105f33a9ac75.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "99d1a9d7e4504a72eb5e3ba343173c10094e3a86c07c1f3bda34f741e854a9cc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3032492077eb1fda02394f801117d82f9848947cedd13df83c155e1d6c9285f7", "generation_seconds_batch": 14.970702097052708, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2758102416992188, "target_z": 0.419921982471305, "ce": 6.173524379730225, "genuineness": 4.18099308013916, "identity": 0.5068414211273193, "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_CART_reg/99d1a9d7e4504a72eb5e3ba343173c10094e3a86c07c1f3bda34f741e854a9cc.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "9a5ee622ee6df6acaa76a25e77f2954e3db07f6a5d34bc5dd8ca85199c7ad2b0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "23f1a59078d036546721ee83e92dab051afd188528642d43ae378f380583f267", "generation_seconds_batch": 20.386883034021594, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1741902828216553, "target_z": 0.34570311547722615, "ce": 5.764476776123047, "genuineness": 4.243753433227539, "identity": 0.6547853350639343, "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_CART_reg/9a5ee622ee6df6acaa76a25e77f2954e3db07f6a5d34bc5dd8ca85199c7ad2b0.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "9a8020bca8b8ea49dd91c4227d27d19c66059947468dfba756346ba70e74c0b3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f753c3779c80f643a66bdd69bbceb6d7b86c7deee4535dd9bb4f4d9cf664fcdb", "generation_seconds_batch": 15.010764572070912, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4280869960784912, "target_z": -0.19921877992871784, "ce": 6.133810997009277, "genuineness": 4.457139015197754, "identity": 0.6347225308418274, "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_CART_reg/9a8020bca8b8ea49dd91c4227d27d19c66059947468dfba756346ba70e74c0b3.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "9b4b332eea369daf408012b809e3904f70a4cc5312a94830772ffc606a2884c5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f1d602097e9dd5b50cf22138157122c17fec2bef5c49ecaaf0c3a9cb0fd8cfc2", "generation_seconds_batch": 18.998006792971864, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7155641317367554, "target_z": 0.010742211306934651, "ce": 6.141753673553467, "genuineness": 4.337893962860107, "identity": 0.3296275734901428, "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_CART_reg/9b4b332eea369daf408012b809e3904f70a4cc5312a94830772ffc606a2884c5.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "9bcd7bc8bf4e4a0afe718dacbbda44a610b01bcd37ce96bb18edbbd264bb030f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "dea662207a2328d0da69b74dc3d6efed5bc5c1345453b233e383aa1273feac7c", "generation_seconds_batch": 14.724796320078894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7488389015197754, "target_z": -0.6953125054415851, "ce": 6.4912309646606445, "genuineness": 4.432034492492676, "identity": 0.3628442883491516, "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_CART_reg/9bcd7bc8bf4e4a0afe718dacbbda44a610b01bcd37ce96bb18edbbd264bb030f.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "a6a1a70dc4dfb7575efd4d7799a9e211739ef21c4af042b58c37545955bf012e", "generation_seconds_batch": 14.60715249297209, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.2674819231033325, "target_z": -1.0468750326495104, "ce": 6.221180438995361, "genuineness": 4.469690799713135, "identity": 0.7577640414237976, "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_CART_reg/9c09e940c5dab53287ffae2d1a8fa5296f41b2df71714656aaa2bb8830bf5162.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "9c778e3f9311b2a458879bb72d083708d0c45615af950c4b6a2ddc01e047ffbb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4d1b2f1fedd957bc7430c36acaa553bc7dca6aa97c2a6fd522c49386bbf289d8", "generation_seconds_batch": 19.758717976976186, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6258255243301392, "target_z": -0.785156209188112, "ce": 6.070269584655762, "genuineness": 4.645420074462891, "identity": 0.6449271440505981, "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_CART_reg/9c778e3f9311b2a458879bb72d083708d0c45615af950c4b6a2ddc01e047ffbb.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "9c7d34bf4a3877b70b93f09517ad38dc7181dd6fd54ce591f283bf57ab9b43ff", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a2c350e614b465adbf2e6a51b9c9e5ccc5f4a54d05a27f576992d75801a0bafb", "generation_seconds_batch": 20.446905766963027, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0002142190933228, "target_z": -0.5117187363960374, "ce": 6.316492557525635, "genuineness": 4.256305694580078, "identity": 0.31162238121032715, "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_CART_reg/9c7d34bf4a3877b70b93f09517ad38dc7181dd6fd54ce591f283bf57ab9b43ff.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "77e7b9294b274370d918c485b2eb5b6fbfeb220e46be2b2892af8ef719e5c207", "generation_seconds_batch": 14.59590975602623, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.566992998123169, "target_z": -0.8281249673504896, "ce": 6.078212261199951, "genuineness": 4.118232727050781, "identity": 0.4858933389186859, "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_CART_reg/9c969db20a6db380e08592f6aed9c9189b3db634d677ebe8f573f593ebdfd2d3.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "9cad7deb7155d1cb88065e757e7db4e902582429e5153c33057f75644dea1633", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "519dbfe1ee6956baa59b905949bbfda597a4bb5a528be81a49c1f007c0bc0076", "generation_seconds_batch": 20.04269958799705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2836799621582031, "target_z": -0.30468749455841493, "ce": 6.213237762451172, "genuineness": 4.1998209953308105, "identity": 0.5478984117507935, "hypothesis": "Die Tür ist noch offen. Ich habe deinen Manteldall 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_CART_reg/9cad7deb7155d1cb88065e757e7db4e902582429e5153c33057f75644dea1633.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "9d3abee205cac98b6fa58c16b814c6d8e6f19b39886b321270c811f564f7648e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2e2bf5f6dad88e79760ac0aab408c829331c58dda3cbd16c994e151d3027269d", "generation_seconds_batch": 19.43283155199606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4187272787094116, "target_z": -0.20605471674851972, "ce": 6.221180438995361, "genuineness": 4.482243061065674, "identity": 0.6441346406936646, "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_CART_reg/9d3abee205cac98b6fa58c16b814c6d8e6f19b39886b321270c811f564f7648e.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "9d892a14aec3c1431d576dbecd307f70924b5fdd4a7b2834d36884fefd5c2ce8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "800ab4d9e91c180c5bb7ea5cc709c87d8a85bed6387fb0fa4c6963ca3ec291f2", "generation_seconds_batch": 14.810796473990194, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8133254051208496, "target_z": 0.8125000435326805, "ce": 5.9431867599487305, "genuineness": 4.005263805389404, "identity": 0.585718035697937, "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_CART_reg/9d892a14aec3c1431d576dbecd307f70924b5fdd4a7b2834d36884fefd5c2ce8.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "9d8f6c0e3a9ac4664fcbf8b41dacfcfb23750b8b6f7625b41d667233503714ca", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9f521e4df68d4f2bfc4b6536ff842ba848d2ed5b373410e5dbbe2a831a80d4fb", "generation_seconds_batch": 19.097396936966106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8332290649414062, "target_z": 0.09667964056632883, "ce": 6.062326908111572, "genuineness": 4.168440818786621, "identity": 0.34166231751441956, "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_CART_reg/9d8f6c0e3a9ac4664fcbf8b41dacfcfb23750b8b6f7625b41d667233503714ca.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "9ddfd4080e5d3371bd3a5f920ddfd98918473099f0dccc68d1e35cfde3b653e0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1417ac4d8c3706ac782254456a3c9b9146f18ab115e83023e685ef015c2f15b7", "generation_seconds_batch": 19.562415135093033, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4603302478790283, "target_z": 0.5546874945584149, "ce": 5.967014789581299, "genuineness": 3.986435890197754, "identity": 0.5240778923034668, "hypothesis": "En concrótora, cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.17647058823529413, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/9ddfd4080e5d3371bd3a5f920ddfd98918473099f0dccc68d1e35cfde3b653e0.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "9dea55ad3160bbe4dc27ed209374d1f2d86081a3aaa96806acc1336473f29a60", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "02c5c24172ac340ddba3c7a7c927e122e58b4887378efd45a7dd2edaf39cf403", "generation_seconds_batch": 21.913951030001044, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.091137170791626, "target_z": -0.44531250544158507, "ce": 6.062326908111572, "genuineness": 4.545003414154053, "identity": 0.48589733242988586, "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_CART_reg/9dea55ad3160bbe4dc27ed209374d1f2d86081a3aaa96806acc1336473f29a60.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": "9e1fde054cf2d73bba0c5752170b4df5be5a296db09e95e7a32faf525d20534c", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "506d41baae117d11240b0b39d860d58305796d4db355f73bc566e395d90e4883", "generation_seconds_batch": 17.116344307083637, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6953548192977905, "target_z": -0.7343749891168299, "ce": 6.324435234069824, "genuineness": 4.519899368286133, "identity": 0.45601513981819153, "hypothesis": "Die Schlüsse liegen auf dem Küchentisch. Ich suchte in der Schublade, dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/9e1fde054cf2d73bba0c5752170b4df5be5a296db09e95e7a32faf525d20534c.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "9ec0f6a71dae53580b44378b63b339738c8eb280d08a97eff5698defbad662cd", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "46bf89316fead11b4b8971f120a995d078dbf886338678c1a62a2c0c412534ad", "generation_seconds_batch": 21.028623520978726, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0257718563079834, "target_z": 0.23730465144949894, "ce": 6.197352409362793, "genuineness": 4.105680465698242, "identity": 0.4820636212825775, "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_CART_reg/9ec0f6a71dae53580b44378b63b339738c8eb280d08a97eff5698defbad662cd.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "9ec9a97c3076ccd924761be8898afc1a2bbb07c9ea4b58fc37cb6fec4ff385f9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "aa9f567a1c717a22d8b1f2a5ed105f3da06284ffc7b46efe8abc10988434a513", "generation_seconds_batch": 21.40882449503988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5457520484924316, "target_z": -0.11328126360396266, "ce": 6.252951145172119, "genuineness": 4.017816066741943, "identity": 0.3552333116531372, "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_CART_reg/9ec9a97c3076ccd924761be8898afc1a2bbb07c9ea4b58fc37cb6fec4ff385f9.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "9edd8aa7369dba41d8ca67927c4959e6e5a9868e0595cb91b4dfd7e858250487", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "bf225f1b20631a8a64310c2099b89b95e22e2fa1b96258268fd4a212cf229040", "generation_seconds_batch": 19.956939787021838, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1767117977142334, "target_z": -0.38281246190890456, "ce": 6.2767791748046875, "genuineness": 4.444586753845215, "identity": 0.4008384346961975, "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_CART_reg/9edd8aa7369dba41d8ca67927c4959e6e5a9868e0595cb91b4dfd7e858250487.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "9f3eb31c1f3f0fb858e33d524a45315fcc2c6bf7b4d44adab51a3261a305eabc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1f1fb5291ca28532c2b5261068e714787538919dec38a71d27ba29bbeab58e5c", "generation_seconds_batch": 19.282007831032388, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3854525089263916, "target_z": 0.5, "ce": 6.173524379730225, "genuineness": 4.174716949462891, "identity": 0.46875566244125366, "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_CART_reg/9f3eb31c1f3f0fb858e33d524a45315fcc2c6bf7b4d44adab51a3261a305eabc.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "9f8473dba16081d5c341102d783eb6438e7ae5a32219b26e297c7026732deb60", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "18ca5bbb858f7d8073735336e8b9149a5cc832995800ecd4268f58b88a548007", "generation_seconds_batch": 19.706848094007, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5245110988616943, "target_z": 0.6015624401425643, "ce": 5.967014789581299, "genuineness": 4.306513786315918, "identity": 0.41677021980285645, "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_CART_reg/9f8473dba16081d5c341102d783eb6438e7ae5a32219b26e297c7026732deb60.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": "9fa9a348fba3a1f26c343895df7ad6943b24208d4b099e835a054bb337d9e4f8", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4a7d7083fbcc0aa2a817592742831db4c5004d9de207f5a4f953ed1bad221899", "generation_seconds_batch": 17.116344307083637, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9199880361557007, "target_z": -0.5703125054415851, "ce": 6.2767791748046875, "genuineness": 4.268857479095459, "identity": 0.4993923008441925, "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_CART_reg/9fa9a348fba3a1f26c343895df7ad6943b24208d4b099e835a054bb337d9e4f8.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "9fd264e4d045b0c3da73203136eea10632695230349f9ee918e84902a8f5fa48", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5f1f736c0bcd5b845b41cbb497fb4f44108f5ff46add29b53cdf02f583aa9e1f", "generation_seconds_batch": 15.997995213023387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3732658624649048, "target_z": -0.23925778869306535, "ce": 6.41180419921875, "genuineness": 4.13078498840332, "identity": 0.3100181519985199, "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_CART_reg/9fd264e4d045b0c3da73203136eea10632695230349f9ee918e84902a8f5fa48.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "a171da6f11419d3305f267200b47c47898c820195845989fee9b38578b867efa", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "8cf627854394aec7479742ba851df2d3eb2bb1da596b453b1dbb0f6795b0bff7", "generation_seconds_batch": 14.97264905704651, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4909307956695557, "target_z": -0.15332035943367117, "ce": 5.935244083404541, "genuineness": 4.15588903427124, "identity": 0.6294580101966858, "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_CART_reg/a171da6f11419d3305f267200b47c47898c820195845989fee9b38578b867efa.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "a1b20393142949ed98fc182189581d8b52c5f7cb067096baa281f639174f9534", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d81a372e09c97653549ebf5be5d86ec506ecd010792c16cfd10d2a77ae95efb6", "generation_seconds_batch": 22.73456231795717, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1153578758239746, "target_z": 0.3027344443802096, "ce": 5.959072113037109, "genuineness": 4.13078498840332, "identity": 0.2538262605667114, "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_CART_reg/a1b20393142949ed98fc182189581d8b52c5f7cb067096baa281f639174f9534.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": "a2839c7c0742bc357080e01499ba2ba13323d37b6262dd0fa72867d65a5662c4", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "17e4547cad4209a279f59d0151ba94940be3fbb5396b357623368128cc7047d4", "generation_seconds_batch": 20.76387565396726, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.377277135848999, "target_z": -0.23632812636039627, "ce": 6.205295085906982, "genuineness": 4.381826400756836, "identity": 0.6234564781188965, "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_CART_reg/a2839c7c0742bc357080e01499ba2ba13323d37b6262dd0fa72867d65a5662c4.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "a306d0f96928582a130e67c0291f7c13b46bcc2ad5e71950e54988ed71a8bcdb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c3fe969e282c67463f6d94539af6843e20343a1b179331021e951e805092dd5c", "generation_seconds_batch": 19.94459084700793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8813648223876953, "target_z": 0.13183593681980188, "ce": 5.982900142669678, "genuineness": 4.2751336097717285, "identity": 0.10669243335723877, "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_CART_reg/a306d0f96928582a130e67c0291f7c13b46bcc2ad5e71950e54988ed71a8bcdb.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "a3589d503a5c46af209e7b4f6421406a65a7ae6b193a31e95e9836d6c33fe54d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8b59dc8b3d0af1281e9ef443cf78e156ea25ff5fac2b43beb42a454d215f73b6", "generation_seconds_batch": 18.987682837992907, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4013450145721436, "target_z": -0.21874997823365974, "ce": 6.292664527893066, "genuineness": 4.394378662109375, "identity": 0.6075237393379211, "hypothesis": "¡Incontrado! La cajita estaba junto a la ventana vieja. ¿Podemos mirar juntos lo que hay dentro?", "wer": 0.11764705882352941, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/a3589d503a5c46af209e7b4f6421406a65a7ae6b193a31e95e9836d6c33fe54d.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "a43931b0ad7dce866b807fbcdc7a37b96230f46e691d0a14c2088e547c6d94ab", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d93409f5617ec38dfb85820f542bf7bd92f052a7050e77ea3e68da447122b0f1", "generation_seconds_batch": 19.2276308509754, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.337164044380188, "target_z": -0.26562501088317014, "ce": 6.062326908111572, "genuineness": 4.356721878051758, "identity": 0.42751815915107727, "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_CART_reg/a43931b0ad7dce866b807fbcdc7a37b96230f46e691d0a14c2088e547c6d94ab.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "a470cecbda3cd6bf9ef6a3adee3f2c2046ed7bc9bd3690176d08631703076d57", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1265df7bd55f83c1d7d6f2a1915f3af139c1e0c64e643865d40619f97fbed4e5", "generation_seconds_batch": 19.635649631964043, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7703852653503418, "target_z": 0.050781220071282156, "ce": 6.300607204437256, "genuineness": 4.287685871124268, "identity": 0.2964756190776825, "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_CART_reg/a470cecbda3cd6bf9ef6a3adee3f2c2046ed7bc9bd3690176d08631703076d57.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "a4a067251b6948e7973963fc4eb2129a090d101cfc04c883c85d4ce0b24a43a8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "39ad2cc425781825d4dc94f2645ba17beaceea517bb101ed36a666983d0f0b94", "generation_seconds_batch": 14.88069182599429, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5992361307144165, "target_z": -0.07421877992871785, "ce": 6.173524379730225, "genuineness": 3.8671910762786865, "identity": 0.45817434787750244, "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_CART_reg/a4a067251b6948e7973963fc4eb2129a090d101cfc04c883c85d4ce0b24a43a8.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "a55d839f3adcf5c54b2f036200856f2fb4fe255da503dd303d3011edbee0c2a7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "31e8058914a624c6da933af234f4ae3b7d1c3b57b2055d0b3dce906d2c5bfe32", "generation_seconds_batch": 15.678358871024102, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7784079313278198, "target_z": 0.05664063180198133, "ce": 6.173524379730225, "genuineness": 4.055472373962402, "identity": 0.49415481090545654, "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_CART_reg/a55d839f3adcf5c54b2f036200856f2fb4fe255da503dd303d3011edbee0c2a7.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "a59b99f34bd0a6c52c4bb042abe66780a4d578f89d2fc0e49f1ddaaaba370d8a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e1639efa0a457fb733ee7cf85ddd2e651b945b98bf734a03a1af2080f3842658", "generation_seconds_batch": 19.022186764981598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5566015243530273, "target_z": 0.624999912934639, "ce": 6.054384231567383, "genuineness": 4.030367851257324, "identity": 0.23228837549686432, "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_CART_reg/a59b99f34bd0a6c52c4bb042abe66780a4d578f89d2fc0e49f1ddaaaba370d8a.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "a5e0b88fe19c9b2ccb7e693fa04ebb6a06847be82b636a934dd4bf59e1c45335", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "bc9f61a1d406b23dc603a5bde32c7fead1239889ce4b6c1e71c9c1e06b9fffa3", "generation_seconds_batch": 19.90601209597662, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9442086219787598, "target_z": 0.17773435731484855, "ce": 5.824047088623047, "genuineness": 4.13078498840332, "identity": 0.2676980793476105, "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_CART_reg/a5e0b88fe19c9b2ccb7e693fa04ebb6a06847be82b636a934dd4bf59e1c45335.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "a60ce9942436d34b291d99b19547f8af62da58124c72f8cf1c3355d00dbc8e04", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "29bb582c474549b74e835bfc8df1b0f793648e4a79eaeb9c71520c7152b2e776", "generation_seconds_batch": 20.814423169009387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.930837631225586, "target_z": 0.1679687581623776, "ce": 5.990842819213867, "genuineness": 3.936227321624756, "identity": 0.6596793532371521, "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_CART_reg/a60ce9942436d34b291d99b19547f8af62da58124c72f8cf1c3355d00dbc8e04.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": "a65f859c696c88d8276a3f12c0b9a42e540bbfcb2ac6e0e67baa2727eaa30fca", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9368218973357ad1e89f66f8a81d3a4873d7a3d296eefad33fd5cea455bdef73", "generation_seconds_batch": 20.85437680897303, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4855824708938599, "target_z": -0.15722654685544293, "ce": 6.340320587158203, "genuineness": 4.193545341491699, "identity": 0.22009369730949402, "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_CART_reg/a65f859c696c88d8276a3f12c0b9a42e540bbfcb2ac6e0e67baa2727eaa30fca.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "a6b50b0aa06359125251426ee2d2c3ddab19ee89e181024a3a437f9af5f45ae8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ada9226c5fafe57e40072b0d49d820eb6a9fa84b2ed55fc517bac8931ecf2c26", "generation_seconds_batch": 20.231323575950228, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5069761276245117, "target_z": -0.14160153597227282, "ce": 6.165581703186035, "genuineness": 4.143336772918701, "identity": 0.41947072744369507, "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_CART_reg/a6b50b0aa06359125251426ee2d2c3ddab19ee89e181024a3a437f9af5f45ae8.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "a6d9b9604155093817e76cb8b21eebb57323c7e27603ebbc40916854c6d3700a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "73bf8083117bb5d9a82a57147fe1f5c866cac9fb8373ed6aa048b4bb7c51f65d", "generation_seconds_batch": 18.936072392971255, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.337164044380188, "target_z": -0.26562501088317014, "ce": 6.443574905395508, "genuineness": 4.319066047668457, "identity": 0.44250309467315674, "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_CART_reg/a6d9b9604155093817e76cb8b21eebb57323c7e27603ebbc40916854c6d3700a.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "a7190bf0aa6e8d668be5f0284b9f44216f90f1b122e8c964e3882895f2275cad", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "33f1f05c0cf735c0761438fa0de5cd6f3254e2679008651cba88b3d129f3c3fd", "generation_seconds_batch": 21.623023084015585, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.377277135848999, "target_z": -0.23632812636039627, "ce": 6.197352409362793, "genuineness": 4.093128204345703, "identity": 0.6898374557495117, "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_CART_reg/a7190bf0aa6e8d668be5f0284b9f44216f90f1b122e8c964e3882895f2275cad.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "a7797b0a8a4f73ea32933531fee4f8d1d60bf3a801f4471116079fdca187f2eb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "001d063d9cb696c49d05a81f6404acf05dcce41ae49bbf3d0bb418ce9cb6dcd8", "generation_seconds_batch": 19.562415135093033, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.374755620956421, "target_z": 0.4921874510257344, "ce": 5.808161735534668, "genuineness": 4.005263805389404, "identity": 0.3439660668373108, "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_CART_reg/a7797b0a8a4f73ea32933531fee4f8d1d60bf3a801f4471116079fdca187f2eb.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "a79f4eee515a9aa8ad911a8748acb80342889d4ad4770d45b2b7934206b2ca2f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7e0a067d35604ec6e57a965c0b910b6ae688da2b3f3ac31a7966154a546988f8", "generation_seconds_batch": 19.827380368020386, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.770538091659546, "target_z": 0.7812500217663403, "ce": 5.772419452667236, "genuineness": 3.9299514293670654, "identity": 0.40493035316467285, "hypothesis": "The door's still open. I put your coat on the chair Tell me when you are ready", "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_CART_reg/a79f4eee515a9aa8ad911a8748acb80342889d4ad4770d45b2b7934206b2ca2f.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "a8214f4350d2de16c4c9f032598e838ac10d2834e0713462a2b7d44955b135d5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "de6dc24640684caabf9883103145f1b43b2d9548c776f8cd4c3c01d643736644", "generation_seconds_batch": 22.161305128014646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1660149097442627, "target_z": -0.39062501088317014, "ce": 6.221180438995361, "genuineness": 4.3504462242126465, "identity": 0.4451640844345093, "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_CART_reg/a8214f4350d2de16c4c9f032598e838ac10d2834e0713462a2b7d44955b135d5.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "56935743645db4f52990c11192cbd0eae7370f09b4f8e1b2b1e1515791223278", "generation_seconds_batch": 14.61928462004289, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5349024534225464, "target_z": -0.8515625272079254, "ce": 6.09409761428833, "genuineness": 4.068024158477783, "identity": 0.650182843208313, "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_CART_reg/a8bd001b1fcafd2f49b65d73e6ed12567e4c9664909a4d08700c9febfd977540.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "a8d1acb5733e29152aebde233a1de2375a7150da6426251540b385d60027aac0", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5151374243e36fdeccd7bf9fef5cbc8303f7eb84566b3b7fc66f7825bed5b9b0", "generation_seconds_batch": 16.79404130706098, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.123227596282959, "target_z": -0.4218750326495104, "ce": 5.935244083404541, "genuineness": 4.231201171875, "identity": 0.7159332036972046, "hypothesis": "De deur er staat nog een nok open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "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_CART_reg/a8d1acb5733e29152aebde233a1de2375a7150da6426251540b385d60027aac0.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "a9169ed90bf09e46010efaee4c2ef29a34e4e2b3da3bdd86c110b5cfd58c0843", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "92c39817d3cf977a7e134308a294ab82ed5f416060a58e06a3db5539138768b8", "generation_seconds_batch": 20.97785390005447, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.139425754547119, "target_z": 0.32031259250694605, "ce": 5.728734970092773, "genuineness": 4.519899368286133, "identity": 0.5261411070823669, "hypothesis": "Oh ! J'ai trouvé la petite boîte ! Elle était près de 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_CART_reg/a9169ed90bf09e46010efaee4c2ef29a34e4e2b3da3bdd86c110b5cfd58c0843.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "a91f2bb970eb1445301be1d594384dc419ec4f046a366d6926611e75ebc051ec", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e5d30c22637a9c779c5f127856b85afddac5071a52b5834709c1b6f8efd0e55a", "generation_seconds_batch": 19.53446306101978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9936814308166504, "target_z": 0.21386717865742427, "ce": 6.308549880981445, "genuineness": 4.293961524963379, "identity": 0.5175057649612427, "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_CART_reg/a91f2bb970eb1445301be1d594384dc419ec4f046a366d6926611e75ebc051ec.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "bda4d716c75a5df6c84a3b90973b73c4283f103c5f4a4e15bf66f73d540d3da8", "generation_seconds_batch": 16.205504458979703, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4814183712005615, "target_z": -0.8906250108831701, "ce": 6.23706579208374, "genuineness": 4.369274139404297, "identity": 0.8035553693771362, "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_CART_reg/a94b7d5da78ff217f8e5a4f7ae4520900bef0ad764d0e4ee898e69de10c28aea.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": "a96bce337b4b0b8846673f6af7729e92142306862a34cf987978ce3ddf214736", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The road turns left behind the bridge.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "94c526bd3ca0d077bc9abf96a89140e2af23b1a5146276159e59454fd62bac18", "generation_seconds_batch": 16.964236039901152, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9788205623626709, "target_z": -0.5273437472792075, "ce": 6.348263263702393, "genuineness": 4.080576419830322, "identity": 0.4377009868621826, "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_CART_reg/a96bce337b4b0b8846673f6af7729e92142306862a34cf987978ce3ddf214736.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "a9e2f013e2d0485c57f2c291bc45f0aa18d8b48bf498dca7fd8b18d98a455206", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e6fc1fcab60bd3aa3a1011806d7bc2b52c0f3dc2f5f28a694c0ed19611ebd55e", "generation_seconds_batch": 21.712897877092473, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.123380422592163, "target_z": 0.3085937690455477, "ce": 6.09409761428833, "genuineness": 4.507347106933594, "identity": 0.6218545436859131, "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_CART_reg/a9e2f013e2d0485c57f2c291bc45f0aa18d8b48bf498dca7fd8b18d98a455206.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "aa0a7671fb8ef663624bea7730fa70a770fb4f80cfeb825322726525ae351ed4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ff7a2aaea2d3732e93938a623fa67902eee15a3177158aa9ac6b94d32f4afaca", "generation_seconds_batch": 19.788465181947686, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.059352397918701, "target_z": 0.9921876251564564, "ce": 5.1469340324401855, "genuineness": 3.616149425506592, "identity": 0.05660061538219452, "hypothesis": "De deur staat nog open. Ik heb je jachts op de stoel gelegd. Zeg maar wanneer je klaar bent.", "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_CART_reg/aa0a7671fb8ef663624bea7730fa70a770fb4f80cfeb825322726525ae351ed4.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "aa9960570b2001ea115e5ae1c12134cc6ec77e8d72ea6d2a42adadd464b06c9c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b4d8f4e91a8037fa8896ccb5168e6d9940f3d0881a7754984891782457bf54f3", "generation_seconds_batch": 20.658666242961772, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1446212530136108, "target_z": -0.40625002176634023, "ce": 6.252951145172119, "genuineness": 4.595211505889893, "identity": 0.5907278060913086, "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_CART_reg/aa9960570b2001ea115e5ae1c12134cc6ec77e8d72ea6d2a42adadd464b06c9c.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "aaca217b316696e7bab94ca01b2df6158d28cf26284a2f1eaa4393ab4efe87fe", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "b6854174853defcea5e28140fdf56d2cdf530215b86fdb89df9e0847c18f6fcb", "generation_seconds_batch": 23.924444830045104, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6528730392456055, "target_z": 0.6953125054415851, "ce": 6.0305562019348145, "genuineness": 3.8671910762786865, "identity": 0.24510960280895233, "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_CART_reg/aaca217b316696e7bab94ca01b2df6158d28cf26284a2f1eaa4393ab4efe87fe.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "ab41a38d9399672e2d459f3e919dad0f456959c4b21ec636d29459694ff1f74b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "12c1685731ad809a6c42fff1544068ef75c0dc2f0d4de0956f71c25716bed581", "generation_seconds_batch": 21.218505374039523, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7651896476745605, "target_z": 0.7773437472792075, "ce": 5.998785495758057, "genuineness": 4.099404335021973, "identity": 0.42830389738082886, "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_CART_reg/ab41a38d9399672e2d459f3e919dad0f456959c4b21ec636d29459694ff1f74b.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "ab506fee71ff8396967029dc841981a72ea362343a02366aeda19af8e7c6b79f", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "29f14b65e3d2e65dbd00338864fa6be12a69ab292cfeadd6033f14422ad859db", "generation_seconds_batch": 16.10707020701375, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9092912077903748, "target_z": -0.5781250108831701, "ce": 6.332377910614014, "genuineness": 4.281409740447998, "identity": 0.48383066058158875, "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_CART_reg/ab506fee71ff8396967029dc841981a72ea362343a02366aeda19af8e7c6b79f.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "39893091ab7fdaa11892d08a4a4b70d63052a3a7713523d68e9c2a162fccb9e5", "generation_seconds_batch": 15.461042138980702, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5135087966918945, "target_z": -0.8671875380910954, "ce": 6.523001670837402, "genuineness": 4.243753433227539, "identity": 0.44032639265060425, "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_CART_reg/ab7a94299f968552a7b966fa3391724b652d8e20483888a7767196f29a7f4a69.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "abc85886c0891729bf207abb1890d978ccc1aa4ee9bc20769fcd25d4139acf5b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3d7f093fbc9fcc053227f1164bb4247d02ec167c06f075c8b24c643c40b79578", "generation_seconds_batch": 19.827380368020386, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.091442584991455, "target_z": 1.0156249238178092, "ce": 5.621508598327637, "genuineness": 3.6224255561828613, "identity": 0.45788198709487915, "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_CART_reg/abc85886c0891729bf207abb1890d978ccc1aa4ee9bc20769fcd25d4139acf5b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "ac1fe5b61144684498550e017b2dfe6942a320fcef73083325e1f96462875e65", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1cf122d20df20268350dfa3afb987a1d49d92190696d7ba7887be56914c9317f", "generation_seconds_batch": 15.050687537062913, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.478896975517273, "target_z": -0.16210934643167843, "ce": 6.284721851348877, "genuineness": 4.137060642242432, "identity": 0.34582218527793884, "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_CART_reg/ac1fe5b61144684498550e017b2dfe6942a320fcef73083325e1f96462875e65.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "ac3f904c14faca99cca6d23ae7fac907cf964fbc3a4ebc6f5e0a3d0e0e47cb2c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0829959525a2c607c84563c108e7f2d8f752d1e8082ea49106ce758e465f7b4b", "generation_seconds_batch": 20.549985943944193, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7274452447891235, "target_z": -0.7109375163247552, "ce": 6.507116317749023, "genuineness": 4.243753433227539, "identity": 0.4808915853500366, "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_CART_reg/ac3f904c14faca99cca6d23ae7fac907cf964fbc3a4ebc6f5e0a3d0e0e47cb2c.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": "ac47e09cc8a2241e0eaf2271e727845ad99c2ff5bccbb6c05e5b81acc1ca3243", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3e24c7ad7b29a9ce995df5e4bd1fb81cbd727d269662298164a43cd3fedb28b1", "generation_seconds_batch": 15.051788567914627, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6846579313278198, "target_z": -0.7421875380910954, "ce": 6.221180438995361, "genuineness": 4.457139015197754, "identity": 0.4896813929080963, "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_CART_reg/ac47e09cc8a2241e0eaf2271e727845ad99c2ff5bccbb6c05e5b81acc1ca3243.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "aca369fc4c135a49a6c8c5f6f629590144a8cc347564d9b80c83dace002ee703", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "c812ed46935fd46c65ed7e86afd6143dd848d618954e7a6796c689238d17afbc", "generation_seconds_batch": 20.503277458949015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8293704986572266, "target_z": 0.8242186928633568, "ce": 5.274016857147217, "genuineness": 3.71028995513916, "identity": 0.19368323683738708, "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_CART_reg/aca369fc4c135a49a6c8c5f6f629590144a8cc347564d9b80c83dace002ee703.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "adb0f5e31a31cf0a37a6d987dce956db7ca32c888b1065dce7206e5da3bae1ce", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b9aac03c2619076b499038377e3492e95dfb92a88cd43821cdacb8db7649314f", "generation_seconds_batch": 21.86206905997824, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9388601779937744, "target_z": 0.17382808282771575, "ce": 5.748591423034668, "genuineness": 4.18099308013916, "identity": 0.16544634103775024, "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_CART_reg/adb0f5e31a31cf0a37a6d987dce956db7ca32c888b1065dce7206e5da3bae1ce.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "e77d9bf9dfc60d06d255d2d8d8f38870890b0573e867496648d3c211f30fa886", "generation_seconds_batch": 14.60276207793504, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6472190618515015, "target_z": -0.7695312853703029, "ce": 6.260893821716309, "genuineness": 4.293961524963379, "identity": 0.51173996925354, "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_CART_reg/adcc0121d6c11de6953f8243f84ad890832f75283c96342c1db246fddc443438.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "add8b701b3054ae8b55a43bf118d765ba2884b5051f36461a723f217e569f675", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "916d7ec188c8b57f10ac94443a881b7a57c8a3cc7c805eb247cde9e06ad0da46", "generation_seconds_batch": 15.004058933001943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.850764274597168, "target_z": 0.839843790811888, "ce": 5.609594821929932, "genuineness": 3.8671910762786865, "identity": 0.2413116991519928, "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_CART_reg/add8b701b3054ae8b55a43bf118d765ba2884b5051f36461a723f217e569f675.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "ae122c611dc0e2769ef9be305b63c9a38de22717a004a455a4da089f24faee29", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "76d506a9816b6daad7a4038643df741d63bca31ed32fece3df77be9361fd0f14", "generation_seconds_batch": 15.632249096990563, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6981817483901978, "target_z": -0.001953137243566392, "ce": 6.1020402908325195, "genuineness": 4.143336772918701, "identity": 0.6621921062469482, "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_CART_reg/ae122c611dc0e2769ef9be305b63c9a38de22717a004a455a4da089f24faee29.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "c4b0b325d674bc4a65e6f059adac8379d30c7ea9e1469e9b82c68aaf6d4d6855", "generation_seconds_batch": 16.305065630935133, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6739611625671387, "target_z": -0.75, "ce": 6.459460258483887, "genuineness": 4.055472373962402, "identity": 0.16044186055660248, "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_CART_reg/ae3130a5db36c2790a31e23135fde257111c56fa9c5b8061c1f3e51a8cb61258.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "ae3214678b3c8332534054863edb3a81447493c578a05997be860d86b7ae2816", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e3f0c476d6de86501c32ace4eb44aa41fac13d21fcd2727e98819253cfe9fc58", "generation_seconds_batch": 14.946655928972177, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1606664657592773, "target_z": -0.3945312853703029, "ce": 6.046441555023193, "genuineness": 4.632867813110352, "identity": 0.6065147519111633, "hypothesis": "He encontrado, ¿verdad? He encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.17647058823529413, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/ae3214678b3c8332534054863edb3a81447493c578a05997be860d86b7ae2816.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "ae39c1c9c09eeb341758bc0753aebe968ca7855b13c178c98efc74467be36bf3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7ec9e9868ee16a05c0372309669936336423fd043b2a360b2b61c2ded69e3488", "generation_seconds_batch": 19.7953302989481, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0779190063476562, "target_z": 0.2753906971010021, "ce": 6.062326908111572, "genuineness": 4.337893962860107, "identity": 0.6459611654281616, "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_CART_reg/ae39c1c9c09eeb341758bc0753aebe968ca7855b13c178c98efc74467be36bf3.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "ae6895584b8b5c94f32c1909a6fc663a54ba0f4470bad3fe83b06caa173ba39a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5f924788e064d0632b870c26407e0f0d8322651b568bdd403e713b5c6283d66b", "generation_seconds_batch": 20.96266799303703, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5726468563079834, "target_z": 0.6367187363960374, "ce": 5.982900142669678, "genuineness": 4.143336772918701, "identity": 0.5812050104141235, "hypothesis": "Ma porte est encore ouverte! J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt !", "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_CART_reg/ae6895584b8b5c94f32c1909a6fc663a54ba0f4470bad3fe83b06caa173ba39a.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "aea765306563efe444f165d0b4d78a82cb4e5ab003df6f504652d23df5b38b3f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4612ac75cced31faba6a9d09613d0401ad49f160fe02036a6f81d1c8cd267946", "generation_seconds_batch": 20.446905766963027, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5283697843551636, "target_z": -0.1259765250891027, "ce": 6.086154937744141, "genuineness": 4.118232727050781, "identity": 0.42850571870803833, "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_CART_reg/aea765306563efe444f165d0b4d78a82cb4e5ab003df6f504652d23df5b38b3f.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "af69203309b0a6749a0613de94d9577d05c2a37561a5b64b9a0b7a004d6e1329", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "56b4c71a8f815975f67161b6685e51501b6307fbd1d88de7b69abff243435019", "generation_seconds_batch": 22.374731275020167, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.313096284866333, "target_z": -0.28320307194454564, "ce": 6.292664527893066, "genuineness": 4.20609712600708, "identity": 0.373681902885437, "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_CART_reg/af69203309b0a6749a0613de94d9577d05c2a37561a5b64b9a0b7a004d6e1329.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "afbd03f625a11ad92c873e68ccd37728f438d1d46cf80fa940443ff50dce44b0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "eae9b3f76ded9984a303fcc898698a185e868efd0ea9d2df871d294dc93942f1", "generation_seconds_batch": 15.758643896086141, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.454981803894043, "target_z": 0.5507812200712822, "ce": 5.808161735534668, "genuineness": 3.9801597595214844, "identity": 0.1747714877128601, "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_CART_reg/afbd03f625a11ad92c873e68ccd37728f438d1d46cf80fa940443ff50dce44b0.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "afe40e43542f971de2ff240936077929868d778f53ccacd96bbcd953c11f7229", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9c778b69028024058c304df811f9182f69adbba903683f9f21e771d5a9ca0a87", "generation_seconds_batch": 20.308075914974324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4735485315322876, "target_z": -0.1660156209188112, "ce": 6.260893821716309, "genuineness": 4.406930446624756, "identity": 0.7473968267440796, "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_CART_reg/afe40e43542f971de2ff240936077929868d778f53ccacd96bbcd953c11f7229.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "afebdb9eea464790c514490669604c867cac109ea8ece61eaec6bb04e745c57c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8b9b1e211ed1b1c9f6024b910f6537a8b7698f841a2e2e875a9fe13a7c396575", "generation_seconds_batch": 19.71243744494859, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6580686569213867, "target_z": -0.03125002176634025, "ce": 6.141753673553467, "genuineness": 4.469690799713135, "identity": 0.6268357634544373, "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_CART_reg/afebdb9eea464790c514490669604c867cac109ea8ece61eaec6bb04e745c57c.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "b0acbb93d68a0b6ab3ee88798725434061a73d286076ad64da57b2b174940d22", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "31998ea49d836fc0e39e1af706f475fe0fd4d7f26fd9dfde83b4f5f27aee6836", "generation_seconds_batch": 19.864074348006397, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2515895366668701, "target_z": -0.3281249673504896, "ce": 6.459460258483887, "genuineness": 4.3881025314331055, "identity": 0.30392155051231384, "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_CART_reg/b0acbb93d68a0b6ab3ee88798725434061a73d286076ad64da57b2b174940d22.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "b0f9f8ecd844ddd806f64579e097210679d158517d3ace4f48b82658b164c105", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ef3e03d00a58000f8b960c2bcfbdc20833e09752942e6195ad4289c2bc7a0866", "generation_seconds_batch": 14.930745444027707, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.658221483230591, "target_z": 0.6992187799287178, "ce": 5.49045467376709, "genuineness": 3.9550557136535645, "identity": 0.07272224128246307, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude ramen. 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_CART_reg/b0f9f8ecd844ddd806f64579e097210679d158517d3ace4f48b82658b164c105.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "b1e9e85d5feef6c3d4dae202bb0e622d53fe5c24c017a4839120d106c68ac606", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f7d6557c35eeca943f4a6b297b94a7321d737aee0aa20946857973e9bbdf534d", "generation_seconds_batch": 20.308075914974324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6580686569213867, "target_z": -0.03125002176634025, "ce": 6.09409761428833, "genuineness": 4.432034492492676, "identity": 0.7083612084388733, "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_CART_reg/b1e9e85d5feef6c3d4dae202bb0e622d53fe5c24c017a4839120d106c68ac606.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "b21206c871ce83a78337f76cbf673b84a82e57c05cbfc4fd7619eb7c66992be2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "66d6bbbadb3aacb5d0f3e79c2d07ddeba0e1a67e3ee45cca4322d4bb26fdf620", "generation_seconds_batch": 19.08817491203081, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8800277709960938, "target_z": 0.13085941173069918, "ce": 6.348263263702393, "genuineness": 4.15588903427124, "identity": 0.5322673320770264, "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_CART_reg/b21206c871ce83a78337f76cbf673b84a82e57c05cbfc4fd7619eb7c66992be2.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "b219eaecfd6e464270b2e4273f7232c52436b8eb2748a6ffae0e6973ec0287a9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "8e3c633ba8469939f3112d52e9d5705a210720a35210d51cea57ff1c1c649850", "generation_seconds_batch": 14.81353822001256, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7170538902282715, "target_z": 0.7421874510257345, "ce": 5.772419452667236, "genuineness": 4.080576419830322, "identity": 0.44772869348526, "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_CART_reg/b219eaecfd6e464270b2e4273f7232c52436b8eb2748a6ffae0e6973ec0287a9.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": "b23ddb5e555c48c1a7e1a472666a979f776bce0e39b6245d55a6a98465c4184d", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0c50853cd1bf1fb373a955cf6de9e2b0ed6e5c968f4f9f4a1c9a7d024dae35af", "generation_seconds_batch": 20.75274171505589, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.370591640472412, "target_z": -0.24121092593663174, "ce": 6.221180438995361, "genuineness": 4.545003414154053, "identity": 0.7595536708831787, "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_CART_reg/b23ddb5e555c48c1a7e1a472666a979f776bce0e39b6245d55a6a98465c4184d.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": "b276874e0f5339ffce077b4721b2a83b42e68d1051b5231438df8b6041c531a8", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "12034516224df0d017a4f6ecb2a0c50693bb1d40537ca6cadb3c191fcb5a22f4", "generation_seconds_batch": 20.814140757080168, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5069761276245117, "target_z": -0.14160153597227282, "ce": 6.221180438995361, "genuineness": 4.65797233581543, "identity": 0.7238959074020386, "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_CART_reg/b276874e0f5339ffce077b4721b2a83b42e68d1051b5231438df8b6041c531a8.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "b33487bc62c5a84e5064ebe7248e463cae0a3d0e19599309fae0793fccb956fb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e6cd977368150e9d113ccb3733d207e1c33104527dc5a4e6064e3a9cfadfdb0b", "generation_seconds_batch": 25.81117937306408, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0190863609313965, "target_z": 0.23242185187326347, "ce": 6.165581703186035, "genuineness": 4.281409740447998, "identity": 0.4965892732143402, "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_CART_reg/b33487bc62c5a84e5064ebe7248e463cae0a3d0e19599309fae0793fccb956fb.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "b34ed1b62a4e50c8cf408a1da4612a8dd21b4c8ea7150fbc259e6dedd0d36411", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d6a758d89fdf906307f82fee7b8da9f2e3160af50d3485eec144d8768508c508", "generation_seconds_batch": 15.591984582017176, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.088615655899048, "target_z": 0.28320307194454564, "ce": 5.935244083404541, "genuineness": 3.9299514293670654, "identity": 0.4758830666542053, "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_CART_reg/b34ed1b62a4e50c8cf408a1da4612a8dd21b4c8ea7150fbc259e6dedd0d36411.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "b3567a8791f5dd180c5181243008a517139ec08f61ed9627b4da329f9051b8e3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4769ccaba80e5b7df80ed6f700a64fbce086000cc864908bc99833a84f1a84f0", "generation_seconds_batch": 21.452150929020718, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.101986885070801, "target_z": 0.2929688452277386, "ce": 5.4944257736206055, "genuineness": 3.923675537109375, "identity": 0.23828254640102386, "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_CART_reg/b3567a8791f5dd180c5181243008a517139ec08f61ed9627b4da329f9051b8e3.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": "b361ac7f938a204ca5f1f06291315723074002b541af4aa7e066057bdef0e5c0", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Your message arrived after I had left.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fd2819f256d67f0d6aefc6767d4449fe9a34b46387fe7969baa232f90f86b746", "generation_seconds_batch": 15.41560586297419, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.235544204711914, "target_z": -0.339843790811888, "ce": 6.41180419921875, "genuineness": 4.15588903427124, "identity": 0.36731553077697754, "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_CART_reg/b361ac7f938a204ca5f1f06291315723074002b541af4aa7e066057bdef0e5c0.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "be74cc41095d0906cd2abec3d9b38965d674041485f3ff210cd26ea53b8ede20", "generation_seconds_batch": 14.678812328027561, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6097801923751831, "target_z": -0.7968750326495104, "ce": 6.24500846862793, "genuineness": 4.494794845581055, "identity": 0.6593595743179321, "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_CART_reg/b429ad869b4f073af736e4e6510807542fc4f5324b0eb7f096c719d9002f4f71.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": "b4d54388f68082dddc36f95ba159e43934e930d9359ed269debb6328a26060be", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2294a0513aa1c2226aff40e869ec3742a5d32d589d8cb0f24d3f39cd18572959", "generation_seconds_batch": 15.96952764200978, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9788205623626709, "target_z": -0.5273437472792075, "ce": 6.340320587158203, "genuineness": 4.65797233581543, "identity": 0.6500887870788574, "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_CART_reg/b4d54388f68082dddc36f95ba159e43934e930d9359ed269debb6328a26060be.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "b51c5eba1be7dc5df880615e92291f0f53e194092582c02d14e70bfc3c7f4360", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7b27af91dedc2367a291d6011beda62ea840297c45028063a9027684415a586f", "generation_seconds_batch": 19.7953302989481, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8813648223876953, "target_z": 0.13183593681980188, "ce": 6.09409761428833, "genuineness": 4.20609712600708, "identity": 0.6364856958389282, "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_CART_reg/b51c5eba1be7dc5df880615e92291f0f53e194092582c02d14e70bfc3c7f4360.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": "b5d656de498dc3c89b31d6a7ef128d15796232795eb6b38a2a132ab28f9fe2f9", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "75676af2d0188f2f043d290c3b96940a8c7b4cdeb60a41eb16b41f123bfcec13", "generation_seconds_batch": 21.095746322069317, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7274452447891235, "target_z": -0.7109375163247552, "ce": 6.205295085906982, "genuineness": 4.432034492492676, "identity": 0.6787947416305542, "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_CART_reg/b5d656de498dc3c89b31d6a7ef128d15796232795eb6b38a2a132ab28f9fe2f9.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "b664e3b4650260797967fc60ef218dcbf4b8cd2dfe3549cfe8c4c52eecae86fd", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8026137b4f514c4ab729342343ac31fbfa2a349fd489156360a07a2522208ee5", "generation_seconds_batch": 19.347513125045225, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.361384630203247, "target_z": 0.4824218518732635, "ce": 5.855817794799805, "genuineness": 4.055472373962402, "identity": 0.6560447812080383, "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_CART_reg/b664e3b4650260797967fc60ef218dcbf4b8cd2dfe3549cfe8c4c52eecae86fd.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "b6dd951c0dae74534eb8f2c09f62d883988cd4d727cde22d9786dd44d1b5b577", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "782533c61e88ec21d20f27a460337f62b3e9c13c82917dd9487b970e773f8815", "generation_seconds_batch": 16.540819116984494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5002906322479248, "target_z": -0.1464843355485083, "ce": 6.149696350097656, "genuineness": 4.18726921081543, "identity": 0.5640458464622498, "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_CART_reg/b6dd951c0dae74534eb8f2c09f62d883988cd4d727cde22d9786dd44d1b5b577.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "b6f9e0de090288c50606236b203c675a439a9c0730d32ddc9424b8837311992a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4ceed64a91ea306babd44b86049c0bb63b4b474ee268ec1cd348248005458dad", "generation_seconds_batch": 19.2276308509754, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8211952447891235, "target_z": 0.08789065356832158, "ce": 6.0146708488464355, "genuineness": 4.137060642242432, "identity": 0.4604252874851227, "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_CART_reg/b6f9e0de090288c50606236b203c675a439a9c0730d32ddc9424b8837311992a.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": "b790d74d294be6a6309ecea7ede0ef0ac8e05c39897fe07b2867c4e593aee987", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "20197b6981b0f9b0619fcaa52f7c988709dbb235160baa98539b2c2f4966d0d4", "generation_seconds_batch": 20.35923095198814, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7541873455047607, "target_z": -0.6914062309544523, "ce": 6.149696350097656, "genuineness": 4.65797233581543, "identity": 0.5637174844741821, "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_CART_reg/b790d74d294be6a6309ecea7ede0ef0ac8e05c39897fe07b2867c4e593aee987.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "b797fa739d05cae170f0cf1fb156f3a3dab27b47b2dc1906628f8c5616047008", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e088301fe515abea6dbad0a04861da576deb6afda0840a2777efe467c74b059b", "generation_seconds_batch": 15.470006876043044, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3506879806518555, "target_z": 0.47460947702971995, "ce": 5.561938762664795, "genuineness": 3.8169827461242676, "identity": 0.4266285300254822, "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_CART_reg/b797fa739d05cae170f0cf1fb156f3a3dab27b47b2dc1906628f8c5616047008.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "b7dc5060e47c070f72854ce252c5bc5d3cf02f10829c5ec49b0691f9a3136ed7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a788ec067931d1d59769913a178aedea2eb1a6606241c799c8d2933b5bfdac0b", "generation_seconds_batch": 20.63154619303532, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2810057401657104, "target_z": -0.3066406318019813, "ce": 6.252951145172119, "genuineness": 3.9801597595214844, "identity": 0.7162827253341675, "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_CART_reg/b7dc5060e47c070f72854ce252c5bc5d3cf02f10829c5ec49b0691f9a3136ed7.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": "b80f5c0c01e87f2d334ff62dc16dcb92aaffac3925c9c30ee892625c01f8e66a", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eb984f1d303e16b6ac11e848503e57d4fb59e024412aefdf14137eb7ecac432b", "generation_seconds_batch": 16.439308190951124, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.24608826637268066, "target_z": -1.0625000435326806, "ce": 6.260893821716309, "genuineness": 4.444586753845215, "identity": 0.5334144830703735, "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_CART_reg/b80f5c0c01e87f2d334ff62dc16dcb92aaffac3925c9c30ee892625c01f8e66a.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": "b8405a88b6ac822c55099329deb7ab1e6621c30e53e835850cd320f7b3e45731", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The keys are on the kitchen table.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f7bcb9195292b5cd26763dca1bfb920ef5d32fd29e3976fdaee55bcde12a8a07", "generation_seconds_batch": 21.014885128941387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8932459950447083, "target_z": -0.5898437472792075, "ce": 6.2767791748046875, "genuineness": 4.3504462242126465, "identity": 0.5139121413230896, "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_CART_reg/b8405a88b6ac822c55099329deb7ab1e6621c30e53e835850cd320f7b3e45731.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "b89d383af2188493822c8b9eb969555d77c2ebe8fbee265c9a377684af54ce89", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1abc08d8c88751e48424694136b88fa86df446f0951394639b05bf987e470429", "generation_seconds_batch": 15.518017295980826, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.679615020751953, "target_z": 0.714843703746527, "ce": 5.089349746704102, "genuineness": 3.7793264389038086, "identity": 0.25917622447013855, "hypothesis": "¡La puerta sigue abierta! ¡He dejado tu abrido 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_CART_reg/b89d383af2188493822c8b9eb969555d77c2ebe8fbee265c9a377684af54ce89.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "b8edda23daf5518fee4f27d454331ad34a8599038c8024600e87182e30867be9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4bed628a23118a1c22b22ebff31bde5627526927b66a6856519b28e821992f12", "generation_seconds_batch": 21.277858664980158, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7595357298851013, "target_z": -0.6875, "ce": 5.879645347595215, "genuineness": 4.507347106933594, "identity": 0.7127620577812195, "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_CART_reg/b8edda23daf5518fee4f27d454331ad34a8599038c8024600e87182e30867be9.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "b926edea6b9548e937b2856572a55bf6e9468515e6b43b7f750160ac9d219a65", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "eab1b603620e7f711d814b2ef2ca6bf460dc99c93c03fbdbc962cef9c49d00c1", "generation_seconds_batch": 21.64173847506754, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1608192920684814, "target_z": 0.3359375163247552, "ce": 6.165581703186035, "genuineness": 4.306513786315918, "identity": 0.457449734210968, "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_CART_reg/b926edea6b9548e937b2856572a55bf6e9468515e6b43b7f750160ac9d219a65.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "b95653ee25a83a353d20f9c3000e75f275cb8ba6e9c389812d09be950a685271", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ad2451f50b75cfd2cb6ad1c40630b896e706b29a06af0030169126552f7a1748", "generation_seconds_batch": 20.64634263899643, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8130198121070862, "target_z": -0.6484375163247552, "ce": 6.109982967376709, "genuineness": 4.419482707977295, "identity": 0.5738569498062134, "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_CART_reg/b95653ee25a83a353d20f9c3000e75f275cb8ba6e9c389812d09be950a685271.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "b9f6898f03d3d1a34cf4b1234ebd4f051e6c8e81cf387b221322a7eacf3467fa", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "97c3ff622cd4969252a0b2d85bc54d4062dc4dde1ce4ff031681a626f6e1fbda", "generation_seconds_batch": 16.540819116984494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.845110297203064, "target_z": -0.625, "ce": 6.308549880981445, "genuineness": 4.457139015197754, "identity": 0.47565075755119324, "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_CART_reg/b9f6898f03d3d1a34cf4b1234ebd4f051e6c8e81cf387b221322a7eacf3467fa.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "ba523b0fe63ce5265baea0d4fee499e3902250f26c96e0ea2445964eeaaed01b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0e69229e96b344671a9946f6e2f6045f7f0d78df3667b5d5669db285992c32b2", "generation_seconds_batch": 19.724742676015012, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9869959354400635, "target_z": 0.2089843790811888, "ce": 6.221180438995361, "genuineness": 4.482243061065674, "identity": 0.4545624256134033, "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_CART_reg/ba523b0fe63ce5265baea0d4fee499e3902250f26c96e0ea2445964eeaaed01b.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "ba60cb4bbb0c7a0e1022de82893aee6bd3896ebdedb7642d68656fd6a746dc63", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9b14046ba3598bc2d19c02db7b2168cab86682aaa1eeb1cb3a8bf7e32c5fcc06", "generation_seconds_batch": 15.632249096990563, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6045845746994019, "target_z": -0.07031250544158506, "ce": 6.157639026641846, "genuineness": 4.425758361816406, "identity": 0.5378754138946533, "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_CART_reg/ba60cb4bbb0c7a0e1022de82893aee6bd3896ebdedb7642d68656fd6a746dc63.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "085e6c6caff56006e2f91b192a1ff9c1a181acf2782835c48718c25043818742", "generation_seconds_batch": 16.92112905299291, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4921152591705322, "target_z": -0.8828124619089046, "ce": 6.221180438995361, "genuineness": 4.369274139404297, "identity": 0.5271438360214233, "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_CART_reg/ba70e7e208a6b15a9236ccd5a69e0e5fd403df2a937202d690effa2cdc45d9f2.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "baf1135b5594089faba768042e9c3109ba3f00783521b5c9ae943b549f0b23bf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "714f8a014dc9d535ce333899abe05933b65f72ac5fdf59eaca8a2042f1c2f496", "generation_seconds_batch": 19.734027337050065, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8293704986572266, "target_z": 0.8242186928633568, "ce": 5.232317924499512, "genuineness": 3.9173994064331055, "identity": 0.30893200635910034, "hypothesis": "Encontré la cajita. Estaba junto a la ventana vieja. ¿Podemos mirar juntos lo que hay dentro?", "wer": 0.11764705882352941, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/baf1135b5594089faba768042e9c3109ba3f00783521b5c9ae943b549f0b23bf.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "baf653d0a50bc4dd211916c356566539f6931209605735a22770ba9af5e4efa3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a04a8bd131520c6b77d4cb8355b1af1e7b094ccbb585176a665831d2b6a3fdc8", "generation_seconds_batch": 14.81353822001256, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8133254051208496, "target_z": 0.8125000435326805, "ce": 5.824047088623047, "genuineness": 4.093128204345703, "identity": 0.578202486038208, "hypothesis": "La porte est encore ouverte ! J'ai posé ton menton sur la chaise ! Dis-moi, quand tu es prêt !", "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_CART_reg/baf653d0a50bc4dd211916c356566539f6931209605735a22770ba9af5e4efa3.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "bb5188b4b2fa067ff7e82adf57b5c9314748f03e27aafff97c3bf81c3ab4c3ae", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f08bd55c5be9c06845da3ecd32c1ca2eadb2164e8aba1430c4a09b9594943b7e", "generation_seconds_batch": 21.985435421927832, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.235697031021118, "target_z": 0.39062501088317014, "ce": 6.252951145172119, "genuineness": 4.2374773025512695, "identity": 0.512965202331543, "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_CART_reg/bb5188b4b2fa067ff7e82adf57b5c9314748f03e27aafff97c3bf81c3ab4c3ae.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "bbc664131596c638d59ccefb51510fe7fc9501e3f09f4297a6329b3d2204e897", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a9a13da9038d079c9c9453743da12573991767a8bbd49896f74cb58669ddb952", "generation_seconds_batch": 22.501690721022896, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3506879806518555, "target_z": 0.47460947702971995, "ce": 6.205295085906982, "genuineness": 3.923675537109375, "identity": 0.4790506362915039, "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_CART_reg/bbc664131596c638d59ccefb51510fe7fc9501e3f09f4297a6329b3d2204e897.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": "bc416122f18fc775ccd38697b20aee2fb75669df700547ffbd88759cf24e69a6", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I wrote the number on the envelope.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8e434c0b45bc6c87aa6606f347e458ac54cb2f9f76cae9c1eb1e2c41eed73002", "generation_seconds_batch": 20.975421852082945, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2542636394500732, "target_z": -0.3261719171722842, "ce": 6.41180419921875, "genuineness": 4.13078498840332, "identity": 0.34721916913986206, "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_CART_reg/bc416122f18fc775ccd38697b20aee2fb75669df700547ffbd88759cf24e69a6.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": "bc5f910d1f38496ffab5019b9a814bf2aa9bb43c06b87a2144e0ba66b58fe202", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The road turns left behind the bridge.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5df411be885993f76fddbffd11ac27da5ba64798400814e3ece1253edebe72a8", "generation_seconds_batch": 21.164766506990418, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7595357298851013, "target_z": -0.6875, "ce": 6.356205940246582, "genuineness": 3.948779582977295, "identity": 0.47440415620803833, "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_CART_reg/bc5f910d1f38496ffab5019b9a814bf2aa9bb43c06b87a2144e0ba66b58fe202.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "bca7316918f2f3b5c3fbd88e7fb017c0bb62307993f3a927065604a9f44984f4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b4c23bb7f478ee1e989487a962a8281fb3701e533323227702e4711858271b44", "generation_seconds_batch": 15.716723824036308, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.32914137840271, "target_z": -0.2714844226138693, "ce": 6.316492557525635, "genuineness": 4.507347106933594, "identity": 0.6928211450576782, "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_CART_reg/bca7316918f2f3b5c3fbd88e7fb017c0bb62307993f3a927065604a9f44984f4.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "bcdf060738e4a963b0a418e08de227e3c950ddb6565f1151ec3e5458dddd15e5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "480b8cbbcb8e363c47a2f848e6e05a042c8619099a8c2f3f399ab92594cbe9a3", "generation_seconds_batch": 21.38856685499195, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.091442584991455, "target_z": 1.0156249238178092, "ce": 5.569881439208984, "genuineness": 3.873466968536377, "identity": 0.3430291414260864, "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_CART_reg/bcdf060738e4a963b0a418e08de227e3c950ddb6565f1151ec3e5458dddd15e5.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": "bd55d724335241074de67201120010497eca471c06f6106b575dbffa5b896bd4", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "82ef66e4bf3d8600f6d0c572bbc7739fa00ee10f6f1b0009d2ca529b9fe93c4d", "generation_seconds_batch": 15.768150239018723, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9841689467430115, "target_z": -0.5234375163247552, "ce": 6.054384231567383, "genuineness": 4.413206577301025, "identity": 0.6883936524391174, "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_CART_reg/bd55d724335241074de67201120010497eca471c06f6106b575dbffa5b896bd4.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "bd6a4f6ec345ae85ff94ed9de80151b2b62bfcede360f643e93acd477de2f782", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ac11bcbc77bbb39919dfa31b90ba6365b1b6ae3d3f6214397eb672a3610a6620", "generation_seconds_batch": 21.628031324013136, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6846579313278198, "target_z": -0.7421875380910954, "ce": 6.372091293334961, "genuineness": 4.030367851257324, "identity": 0.48482728004455566, "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_CART_reg/bd6a4f6ec345ae85ff94ed9de80151b2b62bfcede360f643e93acd477de2f782.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "be724ae6f8f3a8e742444b8b5c63f372b129565ce3befd24fbaf121552e95988", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "bd9d40897a130ea109433e6f7b1918368429a3d8831ca8a97760e91543d8bc5f", "generation_seconds_batch": 24.66216942097526, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.930990219116211, "target_z": 0.8984373857267137, "ce": 5.839932441711426, "genuineness": 3.923675537109375, "identity": 0.036284707486629486, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/be724ae6f8f3a8e742444b8b5c63f372b129565ce3befd24fbaf121552e95988.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "be7e1fc729bdacdc9dc6757fbca2162e7624359c6647788d379f1c815def69d1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ead664014dda3c0848d4612ac57941d95c88dffc9f5b50d9132d349c1ece8c4b", "generation_seconds_batch": 23.905142180039547, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5925506353378296, "target_z": -0.07910157950495332, "ce": 6.173524379730225, "genuineness": 4.519899368286133, "identity": 0.4091871678829193, "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_CART_reg/be7e1fc729bdacdc9dc6757fbca2162e7624359c6647788d379f1c815def69d1.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "bf26649b9a974d0e2d063798ad4a7533c1c45c9b3c19ad637fd588f90be13635", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f48da71850bd80ee1b26d7509e76eab932d227eacb743865c8719bb21a0cdda2", "generation_seconds_batch": 19.98434087098576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.797127366065979, "target_z": 0.07031250544158506, "ce": 6.213237762451172, "genuineness": 4.3504462242126465, "identity": 0.7130642533302307, "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_CART_reg/bf26649b9a974d0e2d063798ad4a7533c1c45c9b3c19ad637fd588f90be13635.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "bf498a1085d179cbcd194514b71eb6235bc1617679073ea2d49e0b97bec050c7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "60260f712381644b55cd422f881ddbd524c28841b71e1e06d461a15071990dd9", "generation_seconds_batch": 21.406358398031443, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8133254051208496, "target_z": 0.8125000435326805, "ce": 5.903473377227783, "genuineness": 3.9676074981689453, "identity": 0.595615029335022, "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_CART_reg/bf498a1085d179cbcd194514b71eb6235bc1617679073ea2d49e0b97bec050c7.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "bf706e4c0b42d822d56a9f69f77ec250e316bacc33c7f995e07c1339402f8157", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e537b325244d6643069ab656de0ea330d31d186d8956e8300c3956dbdd97fb0f", "generation_seconds_batch": 22.143152975011617, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0379586219787598, "target_z": 0.9765625272079254, "ce": 5.7128496170043945, "genuineness": 3.7165660858154297, "identity": 0.5321146249771118, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when ready", "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_CART_reg/bf706e4c0b42d822d56a9f69f77ec250e316bacc33c7f995e07c1339402f8157.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "c0124e08611392ee8cc337b45b3797f3eb32521c14cc6b9cae81ab605229426f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "460cf8d69dda5354301c591cdc223fd7ec528bf1fdc49e5782b5b62788077881", "generation_seconds_batch": 22.546282201074064, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1556236743927, "target_z": 1.0625000435326806, "ce": 5.911416053771973, "genuineness": 3.9550557136535645, "identity": 0.3997194766998291, "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_CART_reg/c0124e08611392ee8cc337b45b3797f3eb32521c14cc6b9cae81ab605229426f.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "c024ab15085b4f2fd3f92fe98181ab0667206a7df3ea2b414f67923b8d567830", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "769f445a24b76025d284f54b84c351f48e0f184a88d8283c1685f10867d13cd3", "generation_seconds_batch": 19.317320410045795, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1741902828216553, "target_z": 0.34570311547722615, "ce": 5.9431867599487305, "genuineness": 3.9048471450805664, "identity": 0.3438968062400818, "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_CART_reg/c024ab15085b4f2fd3f92fe98181ab0667206a7df3ea2b414f67923b8d567830.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "c057819f883a6c633190cdccb0ed6d8c1d3aadcb10d68677e21023e8ea97ad88", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "48abee7dec08c91a5880d5bf65d9ddf1a4b642794c250b6870000f67c835d521", "generation_seconds_batch": 19.460077126976103, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9627753496170044, "target_z": -0.5390624836752448, "ce": 6.1894097328186035, "genuineness": 4.469690799713135, "identity": 0.4761990010738373, "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_CART_reg/c057819f883a6c633190cdccb0ed6d8c1d3aadcb10d68677e21023e8ea97ad88.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "c117710a632935ba517b0329ca7583f6e881e85e5ee53b794b0fadb23a2a2076", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0d582b9cdb4c360cf2797323c576c1477f7c569b231d3edac8a83e102d246ff6", "generation_seconds_batch": 19.940332753001712, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7302721738815308, "target_z": 0.02148433554850829, "ce": 5.9431867599487305, "genuineness": 4.287685871124268, "identity": 0.5214819312095642, "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_CART_reg/c117710a632935ba517b0329ca7583f6e881e85e5ee53b794b0fadb23a2a2076.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "c1192b8b7bc1ab5ed2c6ae4d7c6a82d92ae00ca0b717befa86cc31dd0db9f56b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "204795cb0cef5dc4e700f46f62b99d604e0d5a5486ce877e6ae1bfef17a14ae5", "generation_seconds_batch": 27.004941086983308, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9495570659637451, "target_z": 0.18164063180198134, "ce": 6.372091293334961, "genuineness": 4.494794845581055, "identity": 0.4806061089038849, "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_CART_reg/c1192b8b7bc1ab5ed2c6ae4d7c6a82d92ae00ca0b717befa86cc31dd0db9f56b.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "c1aadaa318814110d037635ce8ee02a7aaa194139c58598367d110f68a651877", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c7ae919613c180c772f39a36dcc0359b9552899cbd0f39ecf8c2b760b1323886", "generation_seconds_batch": 16.208993748994544, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4869195222854614, "target_z": -0.15625002176634026, "ce": 6.252951145172119, "genuineness": 4.400654315948486, "identity": 0.7691241502761841, "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_CART_reg/c1aadaa318814110d037635ce8ee02a7aaa194139c58598367d110f68a651877.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "c2003f428527090fa8e695f93f1b73943a94b0432ef881aabb166f888101665a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9b268019470c4c0328e859d62352956bbd61f1a11a16d32e95eb4b8c211b2681", "generation_seconds_batch": 16.429197776014917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.626131057739258, "target_z": 0.6757813071366432, "ce": 6.125868320465088, "genuineness": 3.7918784618377686, "identity": 0.18503201007843018, "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_CART_reg/c2003f428527090fa8e695f93f1b73943a94b0432ef881aabb166f888101665a.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "c2023634494c7f488e6abb6759ce46349bf1334ab4f16d4942e32abdf4451454", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "6cebee9134acbe6d873afc200eb4b829b2c645ad8513b19a6d83b88778f783fd", "generation_seconds_batch": 19.982762280968018, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.42289137840271, "target_z": 0.5273437472792075, "ce": 5.895530700683594, "genuineness": 4.042920112609863, "identity": 0.12123363465070724, "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_CART_reg/c2023634494c7f488e6abb6759ce46349bf1334ab4f16d4942e32abdf4451454.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "c20f96a3c74e735d5618e030a1c85ed1c43b93d48db05594861db29f49279bbf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "fda099caacda52e5601ac6d25b0f0a05054c0fba2151cca252fc454b36501748", "generation_seconds_batch": 20.084777973010205, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8882031440734863, "target_z": 0.8671875380910954, "ce": 5.391171455383301, "genuineness": 3.9173994064331055, "identity": 0.5271592736244202, "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_CART_reg/c20f96a3c74e735d5618e030a1c85ed1c43b93d48db05594861db29f49279bbf.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "c310cbd6dc680c17c4df70a64de541fc941c0b619a5c1ff8f9eaea554b35a103", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "4e1a401f23811100a1e689a1e02c310da31a4cdef39ad8369115c14b582648d7", "generation_seconds_batch": 21.777649085968733, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4655258655548096, "target_z": -0.17187503264951037, "ce": 6.308549880981445, "genuineness": 4.331617832183838, "identity": 0.5234785676002502, "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_CART_reg/c310cbd6dc680c17c4df70a64de541fc941c0b619a5c1ff8f9eaea554b35a103.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "b94557944c09dbe0e52481db951c8e3111e7b82697b9a62e35523243bbaa1dd3", "generation_seconds_batch": 14.59590975602623, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8237166404724121, "target_z": -0.6406250108831701, "ce": 6.046441555023193, "genuineness": 4.193545341491699, "identity": 0.7154128551483154, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigasta la pequeña puerta de madera. Te esperaré al otro lado.", "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_CART_reg/c3977db0634952a7f9e4083e6135fbbc87199d34c7cb8c3338515d9ae38c18f4.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "c3ae4fc838ae042df2da57c2069f452aa29080f35158960e49c9dd24c94af627", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bb979a5ea488af1c791abfcfe09f8df5c45bf725412f058f9d04ee8ad25986e6", "generation_seconds_batch": 15.518017295980826, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.219804525375366, "target_z": 1.1093749891168299, "ce": 5.645336627960205, "genuineness": 3.9173994064331055, "identity": 0.43318110704421997, "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_CART_reg/c3ae4fc838ae042df2da57c2069f452aa29080f35158960e49c9dd24c94af627.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "c3fe2b344826a17d79964f235789130eb18df45cebe4e419c883550e3f780d98", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3283ac98caa553846feb3e5b270c4dc271e565c27d5f143a06474bf2673e0733", "generation_seconds_batch": 19.097396936966106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8933987617492676, "target_z": 0.14062501088317012, "ce": 5.95112943649292, "genuineness": 4.13078498840332, "identity": 0.1692764163017273, "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_CART_reg/c3fe2b344826a17d79964f235789130eb18df45cebe4e419c883550e3f780d98.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "c41ba3b7387ef104032ad71b9f4aba1c320bfd2c44ffbe90a9dab91dc5b4cd00", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "bbdcd96327f12e1b06d67994ba8f9077e65450892239c5c220cb9c9414ba3b35", "generation_seconds_batch": 21.777649085968733, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5135087966918945, "target_z": -0.8671875380910954, "ce": 6.483288764953613, "genuineness": 4.118232727050781, "identity": 0.32544395327568054, "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_CART_reg/c41ba3b7387ef104032ad71b9f4aba1c320bfd2c44ffbe90a9dab91dc5b4cd00.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "c444ef216f17ce5ae2ae9fc6a36ee0b3419063139666e5b02fe6e04d44543678", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "672ab63bc8ccdab48e590a319a39ace6393afed0f94009b8f1edf22d173bc32a", "generation_seconds_batch": 19.8398582249647, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.59938907623291, "target_z": 0.6562501088317013, "ce": 5.927301406860352, "genuineness": 4.011539936065674, "identity": 0.3596841096878052, "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_CART_reg/c444ef216f17ce5ae2ae9fc6a36ee0b3419063139666e5b02fe6e04d44543678.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "27473ef35d374d6ed5ab2683d98d8f6dc808eb7dfbef770bafec70bcaabdcd48", "generation_seconds_batch": 14.60715249297209, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4921152591705322, "target_z": -0.8828124619089046, "ce": 6.197352409362793, "genuineness": 4.394378662109375, "identity": 0.6360505819320679, "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_CART_reg/c493619abd8899c9b3ce96f2a00a6bcdd900e874ef21436707b986365a8437ee.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "c49d3f164e7f7def456db16253e49b8ee5a2cbf4f5d389e83c15f136bb6c1c02", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b33bc77b22d6d7b384bb3e2cfbff7bd3889e8bde96b4ebebe3918afc80e5189a", "generation_seconds_batch": 17.01352252997458, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.834718942642212, "target_z": 0.8281249673504896, "ce": 5.5539960861206055, "genuineness": 3.7040140628814697, "identity": 0.009068327024579048, "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_CART_reg/c49d3f164e7f7def456db16253e49b8ee5a2cbf4f5d389e83c15f136bb6c1c02.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "c4b0a51fefb4abe08b12c4d8575efe12497f808f6c95f540a27926f8effacb0d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c5809bdbfe64751f770126e1229d57374bd1bd58e143f0562f6f883095e0c47a", "generation_seconds_batch": 19.739560261019506, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8960729837417603, "target_z": 0.14257814812673653, "ce": 6.022613525390625, "genuineness": 4.319066047668457, "identity": 0.4068794548511505, "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_CART_reg/c4b0a51fefb4abe08b12c4d8575efe12497f808f6c95f540a27926f8effacb0d.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "c50b1a9ea1e3dff37f3aea5d70770e7ec33fabb4cba2269689be31c26f30d98b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3efdd65a7af637616bca62a8beae7e533b154e4a56605e642415f35134b4ef8d", "generation_seconds_batch": 14.930745444027707, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6528730392456055, "target_z": 0.6953125054415851, "ce": 5.824047088623047, "genuineness": 3.886019229888916, "identity": 0.24009978771209717, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen eerst 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_CART_reg/c50b1a9ea1e3dff37f3aea5d70770e7ec33fabb4cba2269689be31c26f30d98b.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "c54385315961eb701c7f3f89a661cb99859b750f66ae338595223a8c1371149c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "18cbc384a72fe7b1cb06ddff47026f59f5221d452197f739172b1ad123580e59", "generation_seconds_batch": 21.76622396602761, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6246411800384521, "target_z": -0.055664019647517625, "ce": 6.1020402908325195, "genuineness": 3.653805732727051, "identity": 0.676445722579956, "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_CART_reg/c54385315961eb701c7f3f89a661cb99859b750f66ae338595223a8c1371149c.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "c5595e9e1ae46f639383b01342ac3868ccc00cf8fbfdc817b05e78ab32eb2edc", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5a4ddece1230fcb90730300fc92a2163d7a6998382e8e6b446cff047d375f684", "generation_seconds_batch": 16.24323416291736, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2168247699737549, "target_z": -0.3535156644514917, "ce": 6.133810997009277, "genuineness": 4.281409740447998, "identity": 0.7497765421867371, "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_CART_reg/c5595e9e1ae46f639383b01342ac3868ccc00cf8fbfdc817b05e78ab32eb2edc.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": "c5c100481bbaad71874e3fb862e2ff414b7bc6a4b6c013e0d7d68ee0b184f97a", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cc911d1611d62152e1d8686a46cf12cf704b42e40dff8932a362c23877dc8ec5", "generation_seconds_batch": 16.43756841495633, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5455993413925171, "target_z": -0.8437499782336597, "ce": 6.348263263702393, "genuineness": 4.519899368286133, "identity": 0.40280622243881226, "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_CART_reg/c5c100481bbaad71874e3fb862e2ff414b7bc6a4b6c013e0d7d68ee0b184f97a.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "c5c5b94f89aa64e8cf016c074fbb0a999812fd669891d7ffa597e8af621b4948", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "b98f6f7326eae17d2de55e945dd018da35527dbc326f717bb8654905f89ebbce", "generation_seconds_batch": 21.9035704130074, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7663739919662476, "target_z": 0.04785155773861307, "ce": 5.847874641418457, "genuineness": 4.2249250411987305, "identity": 0.5448887348175049, "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_CART_reg/c5c5b94f89aa64e8cf016c074fbb0a999812fd669891d7ffa597e8af621b4948.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "c62ab298ac6b7f7ab9b4304e1fd222e3cd6ac8aa4c461d7a24607d9144e346ba", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "20951a7f783f9eb9bd8d0bbf574c7a433117439eb4ef333bc0a86c701a2d6856", "generation_seconds_batch": 15.708798031904735, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3373167514801025, "target_z": 0.46484370374652695, "ce": 5.911416053771973, "genuineness": 4.143336772918701, "identity": 0.6151795387268066, "hypothesis": "La porte ! Encore ouverte ! J'ai posé ton manteau sur la chaise ! Dis-moi quand tu es prêt.", "wer": 0.05555555555555555, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/c62ab298ac6b7f7ab9b4304e1fd222e3cd6ac8aa4c461d7a24607d9144e346ba.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "c62e8e66f3954901d858f7b3e36eec81bb07fb9b70e809b596aa943c9314ec71", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b0c4ebf39961c5aade7bcd91cb5fa1be0b40b86362e79f0bdc647276bbd7e51a", "generation_seconds_batch": 25.507614828064106, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.866809368133545, "target_z": 0.8515624401425643, "ce": 5.855817794799805, "genuineness": 3.6977379322052, "identity": 0.12207114696502686, "hypothesis": "Hätte ich meine kleine Schachtel gefunden! Sie stand neben dem alten Fenster. Wir können gemeinsam hineinschauen!", "wer": 0.1875, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/c62e8e66f3954901d858f7b3e36eec81bb07fb9b70e809b596aa943c9314ec71.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "c657c0d80162b5e9c29be4921489c5ebe60355135cd487ea1a8c2b9fd3223c0e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fbc9d0fea9acf2cd6896847dc28c29ad86098e5e67aa8ca5c8e7124b56c2525a", "generation_seconds_batch": 19.60315748094581, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.589876413345337, "target_z": -0.08105471674851972, "ce": 6.0305562019348145, "genuineness": 4.030367851257324, "identity": 0.6492557525634766, "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_CART_reg/c657c0d80162b5e9c29be4921489c5ebe60355135cd487ea1a8c2b9fd3223c0e.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "c6875e1e7e8f3fcbbb0aa7bb62fc686edf4f06b0974bb434fccf33945d58cc4f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "640bcf5d3d5826c9017a8686514bf14158f3dfed5d93ff6bd487300f074e6a87", "generation_seconds_batch": 19.0001315230038, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.551100492477417, "target_z": -0.10937498911682987, "ce": 6.09409761428833, "genuineness": 4.469690799713135, "identity": 0.46668508648872375, "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_CART_reg/c6875e1e7e8f3fcbbb0aa7bb62fc686edf4f06b0974bb434fccf33945d58cc4f.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": "c6b26079518ff758382bdfc9298e1d8c803141b3be7880622b482f27d88783f9", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4f8bd59a6a717b4474e7f1906454291700837bd5397e1bacdc33ae8a5ed6236c", "generation_seconds_batch": 16.157840647967532, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9734721183776855, "target_z": -0.5312500217663403, "ce": 6.24500846862793, "genuineness": 4.582659721374512, "identity": 0.6457051634788513, "hypothesis": "Pon message est arrivé après mon départ. 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_CART_reg/c6b26079518ff758382bdfc9298e1d8c803141b3be7880622b482f27d88783f9.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "c6fcf4454df4d8e6dc8ce75395313aa9cb5f2a141d691d5d2f00894584ac6dda", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "69f15605c4475ca57d720f2d321ead59df5f597d3d8e7b5eb402068bfecfd543", "generation_seconds_batch": 18.87684214406181, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4120417833328247, "target_z": -0.2109375163247552, "ce": 6.1020402908325195, "genuineness": 3.9801597595214844, "identity": 0.47311222553253174, "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_CART_reg/c6fcf4454df4d8e6dc8ce75395313aa9cb5f2a141d691d5d2f00894584ac6dda.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": "c72359047ac4cce86fb4c041d9125e13a246bdc1e1103fdd454b947809ec356c", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Someone moved the little blue chair.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "319a5cb62848fccfc0f395e9de773b1d693d3f8d686d707d928ec1457bc28dae", "generation_seconds_batch": 15.342101266956888, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.968123733997345, "target_z": -0.5351562527207925, "ce": 6.372091293334961, "genuineness": 4.281409740447998, "identity": 0.38825586438179016, "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_CART_reg/c72359047ac4cce86fb4c041d9125e13a246bdc1e1103fdd454b947809ec356c.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": "c749970dab27a5d6f7a3654caa5e2db5ae56415b3df5b861a04f7c3d70be2af1", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "40dae002b05a808f30b584bb15e17f17e9b0d14a0de7f840998f0af4976841fa", "generation_seconds_batch": 14.846603980055079, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1633408069610596, "target_z": -0.3925780610613755, "ce": 6.419747352600098, "genuineness": 4.482243061065674, "identity": 0.5063777565956116, "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_CART_reg/c749970dab27a5d6f7a3654caa5e2db5ae56415b3df5b861a04f7c3d70be2af1.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "c76d66e241773247fb3d5bc84a3e9ca1a538f430f0c9aef40fd819c7a128e589", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3acd9a0f51df4509a93ea4b58a689591b40e5ced4629fd58818716b8da206b9d", "generation_seconds_batch": 19.581233522039838, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8379089832305908, "target_z": 0.10009765250891027, "ce": 6.141753673553467, "genuineness": 4.030367851257324, "identity": 0.2769660949707031, "hypothesis": "The door still open. I 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_CART_reg/c76d66e241773247fb3d5bc84a3e9ca1a538f430f0c9aef40fd819c7a128e589.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "c787c891e3baedfbeb03a33143fdba9da25b326d0e0ca437a21c981b369a3751", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4fd1494e4eee95e5310d4e968540dc76794ac2362b3f92e9497e1dea247551c9", "generation_seconds_batch": 22.374731275020167, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4468064308166504, "target_z": -0.1855469062891141, "ce": 6.141753673553467, "genuineness": 4.331617832183838, "identity": 0.16265490651130676, "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_CART_reg/c787c891e3baedfbeb03a33143fdba9da25b326d0e0ca437a21c981b369a3751.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "c78dd65ca758728ee4eb8ef77fe31fa5f1a6feca1968e1359bee4a6944f94444", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2bd19e0b9d8b3b645dedb9c2410710615e8c1650a6ad7e2fc38a886b44fc7491", "generation_seconds_batch": 25.032887223991565, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3480136394500732, "target_z": 0.47265625272079254, "ce": 5.673136234283447, "genuineness": 4.13078498840332, "identity": 0.26851487159729004, "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_CART_reg/c78dd65ca758728ee4eb8ef77fe31fa5f1a6feca1968e1359bee4a6944f94444.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "c7cea433f3c18658ce3d18ab9830a68f2998cfaaa06ab97a11a843c13f591f91", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4eeb5d5760ee33960c32fbd8d3eb8a848726b2f9c9a00ad374263abed6d606a6", "generation_seconds_batch": 19.427252536988817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9575796127319336, "target_z": 0.18749995646731948, "ce": 6.006728172302246, "genuineness": 3.7667744159698486, "identity": 0.30193889141082764, "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_CART_reg/c7cea433f3c18658ce3d18ab9830a68f2998cfaaa06ab97a11a843c13f591f91.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "c81cae7d1e63aaf8ca8bc583b92a9e42130007069cf9dccc02265fa812900527", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2bea8a68a82de982c4e0de88468204e6cf7909bc99b650b1491126c42614e4fb", "generation_seconds_batch": 19.460077126976103, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2863541841506958, "target_z": -0.30273435731484855, "ce": 6.23706579208374, "genuineness": 4.507347106933594, "identity": 0.7251632809638977, "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_CART_reg/c81cae7d1e63aaf8ca8bc583b92a9e42130007069cf9dccc02265fa812900527.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "5e045584fe2f2a889edd7ee00382d4d3ade30e95839586b668cd3953b81006dd", "generation_seconds_batch": 14.60276207793504, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6151286363601685, "target_z": -0.7929687581623776, "ce": 6.23706579208374, "genuineness": 4.344170093536377, "identity": 0.7062758207321167, "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_CART_reg/c82131bfaf89850d394bf3a21117deb94436b65306054c3e1cbac0c65e2f913d.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "c84156a7f71fd7c505b46ff6fd8451eddafef7a940caa6ef729aa37c5880bd0d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "db4d015abd6a2d19a75c2a864d7bede39f0cd2c24588a71293ae08f4b1020d02", "generation_seconds_batch": 21.868402069085278, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7449803352355957, "target_z": 0.03222654685544294, "ce": 6.165581703186035, "genuineness": 4.306513786315918, "identity": 0.721401035785675, "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_CART_reg/c84156a7f71fd7c505b46ff6fd8451eddafef7a940caa6ef729aa37c5880bd0d.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": "c86dba3ff20fba150cd6467640f9509f56bed8253b26908de0978b85fee0106c", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2880368d443310dc97710e9e9b560cfe30cfb071de3389f8a208d8ac8506e688", "generation_seconds_batch": 15.188815360073932, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8023229837417603, "target_z": -0.6562500217663403, "ce": 6.086154937744141, "genuineness": 4.319066047668457, "identity": 0.638166069984436, "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_CART_reg/c86dba3ff20fba150cd6467640f9509f56bed8253b26908de0978b85fee0106c.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "c884c943c738a43febed2662a303ded8e16150f9ae6255c41726c0be578f3920", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "cb86c87b2004efede95762673863c4ff9d29a8d52b534d2f9a23247e2a3f3f85", "generation_seconds_batch": 15.429269929067232, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5726468563079834, "target_z": 0.6367187363960374, "ce": 5.561938762664795, "genuineness": 4.042920112609863, "identity": 0.5914922952651978, "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_CART_reg/c884c943c738a43febed2662a303ded8e16150f9ae6255c41726c0be578f3920.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "c3335ab9231cb04ee908d49dfe7e4d912c838601cac849bda7a2c2385d90a599", "generation_seconds_batch": 16.4097449160181, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.32096612453460693, "target_z": -1.0078124619089046, "ce": 6.523001670837402, "genuineness": 4.519899368286133, "identity": 0.34687215089797974, "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_CART_reg/c8907269f81c0a33f09ba3b9d9f76b6acb2a8250e6881dbc11786cea01719e9d.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "c8939b826fe6b8204d8b184690db8087c5a6734a29a885c56f5f0e03f1f250b2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "44cdccbe2a769925e45d8808b46681dc8031dc0385520f03c9bb5173a9f35f49", "generation_seconds_batch": 16.240866589942016, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7209124565124512, "target_z": 0.014648398728706425, "ce": 6.165581703186035, "genuineness": 4.268857479095459, "identity": 0.6813921928405762, "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_CART_reg/c8939b826fe6b8204d8b184690db8087c5a6734a29a885c56f5f0e03f1f250b2.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "c8d82b27de75b5ea8bff088572fc5c2c4caa37d13a1a1ae77732ae96768913a0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "1e852357cd19c00a03a054b437c2a5cf0309467fd068103fb5b7b86f5e7e9e4e", "generation_seconds_batch": 16.328598621999845, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3025522232055664, "target_z": 0.4394531807762469, "ce": 5.839932441711426, "genuineness": 4.13078498840332, "identity": 0.25850456953048706, "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_CART_reg/c8d82b27de75b5ea8bff088572fc5c2c4caa37d13a1a1ae77732ae96768913a0.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "c8e075472614764bb80ed6eda8a8d4764e670dce9716030cfe91bfe696e5e2c2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "78436df98fda8da3304257416582770267ebb67f51842f01eca0b47c79c460da", "generation_seconds_batch": 23.905142180039547, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8345662355422974, "target_z": 0.09765625272079254, "ce": 5.879645347595215, "genuineness": 4.394378662109375, "identity": 0.20298919081687927, "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_CART_reg/c8e075472614764bb80ed6eda8a8d4764e670dce9716030cfe91bfe696e5e2c2.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "c9d09a5ccd0d7935c24e200a92e9a94586e9a3a92cac6c0359fd3a6a51111800", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f5e6dedc735edd8bd870736a8f9e2aad4c3468ec26a40c8f1d4ab4915e284c64", "generation_seconds_batch": 15.293078011018224, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.567298412322998, "target_z": 0.6328124619089046, "ce": 6.038498878479004, "genuineness": 4.15588903427124, "identity": 0.020867133513092995, "hypothesis": "De deur staat nog open? Ik heb je jas op de stoel gelegd. Zie zeg maar wanneer je klaar bent.", "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_CART_reg/c9d09a5ccd0d7935c24e200a92e9a94586e9a3a92cac6c0359fd3a6a51111800.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "ca34c213f9d19a42dcb1e668622703bbf61f3efa8ab49dfd1e7dba054297fc6c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "39319ea273f063838140b604ae5d5a45b09b66c5e47e849c213ae43d7950d626", "generation_seconds_batch": 23.27186566707678, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3854525089263916, "target_z": 0.5, "ce": 5.637393951416016, "genuineness": 4.162164688110352, "identity": 0.6012348532676697, "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_CART_reg/ca34c213f9d19a42dcb1e668622703bbf61f3efa8ab49dfd1e7dba054297fc6c.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "ca4f82f6de785e046a00c3cc7fcd88e9403db4e505c30743e1d670e4c3f33b18", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "47936feef5c925333c4773cee5b4d6046048aa72d40354d63270c3b6b8b29de1", "generation_seconds_batch": 19.24527993099764, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5245110988616943, "target_z": 0.6015624401425643, "ce": 5.7406487464904785, "genuineness": 3.948779582977295, "identity": 0.43434083461761475, "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_CART_reg/ca4f82f6de785e046a00c3cc7fcd88e9403db4e505c30743e1d670e4c3f33b18.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "cab1f7e95d2c5b4d7de5349b2e4fa1208b29fdf791062fe5fe0d0b012b837c1e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c5fbdd9f87d0db33f23ffa13ae7021a8254a38c489976ed253c3d305af8dca4d", "generation_seconds_batch": 21.583754868945107, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.791931629180908, "target_z": 0.7968749455841494, "ce": 5.788304805755615, "genuineness": 3.9048471450805664, "identity": 0.6194461584091187, "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_CART_reg/cab1f7e95d2c5b4d7de5349b2e4fa1208b29fdf791062fe5fe0d0b012b837c1e.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "cb6e75a9504e4322b6c0e7a88c54453023711d7d4cd0bcbb9c8e9af94376c311", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "49f42be526cf7b5ead12b4b1b79d368cada6a32c8dcd5060d56232e952d24d89", "generation_seconds_batch": 16.06865647097584, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4121944904327393, "target_z": 0.5195311983049419, "ce": 5.919358730316162, "genuineness": 4.036643981933594, "identity": 0.433014839887619, "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_CART_reg/cb6e75a9504e4322b6c0e7a88c54453023711d7d4cd0bcbb9c8e9af94376c311.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "cb7d6953d2de429ca5513818510627686b09ea4065c58aecd18b0aa83e157bd6", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e09b0b13d8cd38cda8c4129f1c78f4feb317635f150fe685ab41f2b138e8b9ca", "generation_seconds_batch": 21.301889894064516, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.706357002258301, "target_z": 0.7343749020514688, "ce": 5.617537498474121, "genuineness": 4.111956596374512, "identity": 0.13286492228507996, "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_CART_reg/cb7d6953d2de429ca5513818510627686b09ea4065c58aecd18b0aa83e157bd6.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "cd50a977ba6c90a4743c1d7fb866e6c131295ad111eda078fae8ed80c3a6b7a2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "440a524fffc79569ebfc21f49997207258ca652a611453c36754145ae56e61e0", "generation_seconds_batch": 20.206531993928365, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2408926486968994, "target_z": -0.3359375163247552, "ce": 6.205295085906982, "genuineness": 3.9676074981689453, "identity": 0.2955039441585541, "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_CART_reg/cd50a977ba6c90a4743c1d7fb866e6c131295ad111eda078fae8ed80c3a6b7a2.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "cd9bca84fdbc09615339c13a9aec36769ec1d4359a1923f37a22b1799786eff7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "0d3a6675841623c35e5c53f266d30175a8a021d444948d638432815a24db349e", "generation_seconds_batch": 19.707248407998122, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.738447666168213, "target_z": 0.7578125489742655, "ce": 5.186647415161133, "genuineness": 3.7416701316833496, "identity": 0.3689541220664978, "hypothesis": "¡La puerta sigue abierta! He dejado tu abrijo sobre la silla. Dime cuándo estás listo.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/cd9bca84fdbc09615339c13a9aec36769ec1d4359a1923f37a22b1799786eff7.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "cdafba2ccb8b37be51c1137f2f3cffe11574c73f902381cee0529fda4b4d14a1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "436b0e5872b20afff78253edb03d044913b50a589d36b9f74888ffe80ce01548", "generation_seconds_batch": 16.06865647097584, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.834718942642212, "target_z": 0.8281249673504896, "ce": 5.6334228515625, "genuineness": 4.300237655639648, "identity": 0.6089401245117188, "hypothesis": "¡Eh! ¡He encontrado la cajita! Estaba junto a la ventana! ¡Vieja! Podemos mirar juntos lo que hay dentro!", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/cdafba2ccb8b37be51c1137f2f3cffe11574c73f902381cee0529fda4b4d14a1.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "cf0cfee2c868edd971628f8881156c4d415fd67c9ff214d45d3a16d333976658", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "70a6ecd7121b7e2718f31cb2e7acee5a2040013cff540fef467857d2722fb94b", "generation_seconds_batch": 15.004058933001943, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.759841203689575, "target_z": 0.7734374727920746, "ce": 5.665193557739258, "genuineness": 4.011539936065674, "identity": 0.5211448669433594, "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_CART_reg/cf0cfee2c868edd971628f8881156c4d415fd67c9ff214d45d3a16d333976658.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "8159211749f586876e87efe6cedbd24b845634582e12c13e27d7f7b6c30d841b", "generation_seconds_batch": 16.59513765689917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4172375202178955, "target_z": -0.9374999564673195, "ce": 6.324435234069824, "genuineness": 4.557555198669434, "identity": 0.613798975944519, "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_CART_reg/cf68e6506caaaad7e7371e72de9d5417fef89211300487ff83acd2484eaab510.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "cf9deae27708f5da502ef89d518cf9f88baccff4e3bb60451ec154349dd739c1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b24391928dc8238c135bcc2b6fb60dbacea4d6a70b0610f087bbc7552e5a2d1e", "generation_seconds_batch": 24.037589969928376, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.471027135848999, "target_z": 0.5625000435326805, "ce": 5.927301406860352, "genuineness": 4.061748504638672, "identity": 0.6215969324111938, "hypothesis": "Ma porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt.", "wer": 0.05555555555555555, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/cf9deae27708f5da502ef89d518cf9f88baccff4e3bb60451ec154349dd739c1.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "cfa91765749435416ec882d759ecb95fc794198702f40b045369ffd88a15a6eb", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "42a74376059d4dd4856b6535b7d1f8d54d8503c33dbf06300e4230893cbff145", "generation_seconds_batch": 20.573735035024583, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7476545572280884, "target_z": 0.03417968409900934, "ce": 6.022613525390625, "genuineness": 4.080576419830322, "identity": 0.47876858711242676, "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_CART_reg/cfa91765749435416ec882d759ecb95fc794198702f40b045369ffd88a15a6eb.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": "cfe9298a2bc8fbf2a1541d2c9cdd4906808c6911dc0a267a3cf6bb1e1d76e607", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Someone moved the little blue chair.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9f527f5cefa00550fb68de4a2744b968eed9096152584550f6ef4426bb305df6", "generation_seconds_batch": 15.342101266956888, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9360332489013672, "target_z": -0.5585937690455477, "ce": 6.284721851348877, "genuineness": 4.168440818786621, "identity": 0.38237130641937256, "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_CART_reg/cfe9298a2bc8fbf2a1541d2c9cdd4906808c6911dc0a267a3cf6bb1e1d76e607.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": "d09b7b78d437df51dc91004a5b908db6617cc1f144a74ac47f521d496883be81", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d2aefb420b3d9181ed0c965495750a6d2e9423c39fb78d4b660e12d6b22a3652", "generation_seconds_batch": 15.051788567914627, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9146396517753601, "target_z": -0.5742187363960374, "ce": 5.974957466125488, "genuineness": 4.080576419830322, "identity": 0.4935877025127411, "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_CART_reg/d09b7b78d437df51dc91004a5b908db6617cc1f144a74ac47f521d496883be81.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "3ceb45dad5aede606a2f08748fb6b0344e5976a23c0442cdc35642eeeab40937", "generation_seconds_batch": 16.05141171696596, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.34235966205596924, "target_z": -0.9921875380910954, "ce": 6.41180419921875, "genuineness": 4.105680465698242, "identity": 0.2054443359375, "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_CART_reg/d11caac31fa3732e2517803823f1b039213fb31e9438619fd83d8b232c894b7d.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "d11f95b00105241336d36759b0410a755d766a508c259d2e15ab92eec04a438b", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5a5f6b5cfbac99f30e6179bc311f5ae0af0ab5ba73207c3b1951632d3094c6bf", "generation_seconds_batch": 16.79404130706098, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.16884183883667, "target_z": 0.34179684099009333, "ce": 6.022613525390625, "genuineness": 4.011539936065674, "identity": 0.6753402948379517, "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_CART_reg/d11f95b00105241336d36759b0410a755d766a508c259d2e15ab92eec04a438b.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "d176acc51c8729a8394e184cf5b925cd511d0f8beee6fac85ff65080211a0074", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "65e96a3936b26e3287952c192bf42ebb3c2da5d125dd862afe856aa0080a09b3", "generation_seconds_batch": 20.433072619955055, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6981817483901978, "target_z": -0.001953137243566392, "ce": 6.133810997009277, "genuineness": 4.231201171875, "identity": 0.7015018463134766, "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_CART_reg/d176acc51c8729a8394e184cf5b925cd511d0f8beee6fac85ff65080211a0074.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "085fc11f8260bf7a0cbb87e9d2f38b25e72f49d7e6ef955a3484a956abde77de", "generation_seconds_batch": 14.59807966905646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.599083423614502, "target_z": -0.8046874945584149, "ce": 6.260893821716309, "genuineness": 4.105680465698242, "identity": 0.6897587776184082, "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_CART_reg/d2319f0ba4d4a13f099c61fd2524d2cd47bb09bca092815ed8a29e7f5b0c527c.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": "d28aa2167804ec85bc49867c054c4591bb8ce248a1230756268e98809150beec", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Your message arrived after I had left.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6a58f172bd346a974dc9af4f1af18f2a3a5c18544fd5f2bab0624f390571ef25", "generation_seconds_batch": 15.41560586297419, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0002142190933228, "target_z": -0.5117187363960374, "ce": 6.395919322967529, "genuineness": 4.024092197418213, "identity": 0.248144268989563, "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_CART_reg/d28aa2167804ec85bc49867c054c4591bb8ce248a1230756268e98809150beec.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "d2a3a89e88c74f6e0863fa4906a3dae54c12da686fa84f95a54a9a028821cfd9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e3c5f82453b4f7d9a75d48c05731b64bd47ed025c4ca4cd924550bdf36a6b93c", "generation_seconds_batch": 15.128914439002983, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3157703876495361, "target_z": -0.28125002176634023, "ce": 6.38003396987915, "genuineness": 4.494794845581055, "identity": 0.3349751830101013, "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_CART_reg/d2a3a89e88c74f6e0863fa4906a3dae54c12da686fa84f95a54a9a028821cfd9.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "d2bb3abf7d5bf3990690b28171b3f9db490b62f385aa4973c211eff63f68eb18", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "665ce0aa36a8ce667069ab9f9e82b5fe2ac1c5387c9fb5c4d19de097c077ac82", "generation_seconds_batch": 22.9633475280134, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.42793428897857666, "target_z": -0.9296874945584149, "ce": 6.427689552307129, "genuineness": 4.6705241203308105, "identity": 0.4681687355041504, "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_CART_reg/d2bb3abf7d5bf3990690b28171b3f9db490b62f385aa4973c211eff63f68eb18.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "d33f45dbaf26dac441585b73bc544c1a82f5a4102f09c1995920dcc1bc61810f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "e95bd36c17358e92837982e72c2d37f8eb0c035b21fc36a9fb1f5eadf5aa8bd4", "generation_seconds_batch": 20.313484448939562, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.406846046447754, "target_z": 0.5156249238178091, "ce": 5.911416053771973, "genuineness": 4.243753433227539, "identity": 0.731573224067688, "hypothesis": "Mais ce que j'ai trouvé la petite boîte. 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_CART_reg/d33f45dbaf26dac441585b73bc544c1a82f5a4102f09c1995920dcc1bc61810f.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "d394658f86558ce24a7159695893e2ebaa0f97f59224ee45136e5889c866256a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "262b2c560f310e3db612317854057df5824ccfb4c100498432773a8a3779c640", "generation_seconds_batch": 21.80552641209215, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5886921882629395, "target_z": 0.6484375598574357, "ce": 5.510311126708984, "genuineness": 3.9173994064331055, "identity": 0.495152086019516, "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_CART_reg/d394658f86558ce24a7159695893e2ebaa0f97f59224ee45136e5889c866256a.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "f25e7146fbe9aa5a884e6d64bc90ed611190ed3875968cbb073a127b298f1745", "generation_seconds_batch": 14.767348385998048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7167484760284424, "target_z": -0.7187499782336597, "ce": 6.109982967376709, "genuineness": 4.218649387359619, "identity": 0.6265707015991211, "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_CART_reg/d3fab0ead1fc9dcf899f59bddb0c9d4b54373ba1829b6dc49f77fc4027ce1740.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "d53c857294e51894d0c79a65e50856956eb2621c20e3e36aab77ef133b3c85cf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4f8122d525588fb6f75ee8883a344a2697194ce9cf8799e941c52abb83d33d34", "generation_seconds_batch": 16.514050760073587, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4763755798339844, "target_z": 0.5664063180198133, "ce": 5.919358730316162, "genuineness": 4.030367851257324, "identity": 0.40128591656684875, "hypothesis": "I found the little box was beside the old window. We can look inside together.", "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_CART_reg/d53c857294e51894d0c79a65e50856956eb2621c20e3e36aab77ef133b3c85cf.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "d55e4c12a4f466325e471fbde969b5e71de09261523c1afed2cf3d9567d9a03b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7d6768018b512492319c809910e442a9d2f7247fb1a3b0bc54e6d2a5fe46c239", "generation_seconds_batch": 18.98166206595488, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7810821533203125, "target_z": 0.05859376904554772, "ce": 6.205295085906982, "genuineness": 4.118232727050781, "identity": 0.771414577960968, "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_CART_reg/d55e4c12a4f466325e471fbde969b5e71de09261523c1afed2cf3d9567d9a03b.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "d5777fe37c96a7fa28100db0339af811aad961115d2e33650960a1a8f2de6fbf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4e463e8ad52e18f1b9170dd3e78c498ba57a9ba64c90b1438248733cce641c0c", "generation_seconds_batch": 15.429269929067232, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8653196096420288, "target_z": 0.12011720042376453, "ce": 6.038498878479004, "genuineness": 4.494794845581055, "identity": 0.4173615574836731, "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_CART_reg/d5777fe37c96a7fa28100db0339af811aad961115d2e33650960a1a8f2de6fbf.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "d57a34c2a5273037c2c82ea53dc6aeef85d4ae1dbe0c2a7adeb955457e2523ad", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "cb05a9b15a812b1c6bff311b9a21bc6657a26bcabec79e5579c129f27ff16fc6", "generation_seconds_batch": 15.279747102060355, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6689181327819824, "target_z": 0.7070311547722614, "ce": 5.756534099578857, "genuineness": 4.017816066741943, "identity": 0.4379498362541199, "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_CART_reg/d57a34c2a5273037c2c82ea53dc6aeef85d4ae1dbe0c2a7adeb955457e2523ad.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": "d5a19bffeba0e257e9bdba8bdbeb59b65675a0af3ddf84d7d02a2d5ea324baba", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "758ee943f44f8dee26298f6313b97e3848ba714b11583567a108f15a8f901285", "generation_seconds_batch": 15.685390901053324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0938113927841187, "target_z": -0.4433593681980187, "ce": 6.125868320465088, "genuineness": 4.570107460021973, "identity": 0.6944690942764282, "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_CART_reg/d5a19bffeba0e257e9bdba8bdbeb59b65675a0af3ddf84d7d02a2d5ea324baba.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "d5f4829f0e9dc181e7f43d3242be8509260cc0ccedbe605bb5d7e86d9d48a4f1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7e0e4b46fe9599d4da73eabf1763eaa60b1efaafed8f3ef31daed254874c3602", "generation_seconds_batch": 20.035419023944996, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.678125262260437, "target_z": -0.01660153597227282, "ce": 6.38797664642334, "genuineness": 4.507347106933594, "identity": 0.45800504088401794, "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_CART_reg/d5f4829f0e9dc181e7f43d3242be8509260cc0ccedbe605bb5d7e86d9d48a4f1.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "d61935c403c01288e556f4c130bd443b3fc8b80a6cdf52c96cf6d7ef1c87c543", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4e27679478a95dde2ee44c5e35670d48878ad0801cb88b5dac37c61201a932cb", "generation_seconds_batch": 19.008948259055614, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.684810757637024, "target_z": -0.011718736396037342, "ce": 6.157639026641846, "genuineness": 4.444586753845215, "identity": 0.6208685636520386, "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_CART_reg/d61935c403c01288e556f4c130bd443b3fc8b80a6cdf52c96cf6d7ef1c87c543.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "d628ca91da1d4a39eccf0fdee043114afc5a34f07a5e96ee79fde6de48936919", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "75fe89e3aebe913aac737cfe9577828830e0164bfdf4199c81cff12b96c262d3", "generation_seconds_batch": 19.48264504992403, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3373167514801025, "target_z": 0.46484370374652695, "ce": 5.855817794799805, "genuineness": 4.017816066741943, "identity": 0.6209519505500793, "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_CART_reg/d628ca91da1d4a39eccf0fdee043114afc5a34f07a5e96ee79fde6de48936919.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "d63aef129c854ad14bb37a96e0ad7c1b1ce9876f4177614503bf75b41af90eec", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "58cb646917fe61cdc914eb86a2bbe24cf5453b7cd4e1fe01ae85f1a14d9832df", "generation_seconds_batch": 19.827381295035593, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9256420135498047, "target_z": 0.8945312853703029, "ce": 5.982900142669678, "genuineness": 3.835810661315918, "identity": -0.08366180211305618, "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_CART_reg/d63aef129c854ad14bb37a96e0ad7c1b1ce9876f4177614503bf75b41af90eec.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "d6529579678f435b5a1a9ba478d5edc2df4edb452bcf533365c3fc7cde322b04", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "dcc753aa6703c39ce6eac27873a92ce622cd64d503247fb36513d20aaa3767e4", "generation_seconds_batch": 20.24757582996972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3157703876495361, "target_z": -0.28125002176634023, "ce": 6.046441555023193, "genuineness": 4.432034492492676, "identity": 0.6359640955924988, "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_CART_reg/d6529579678f435b5a1a9ba478d5edc2df4edb452bcf533365c3fc7cde322b04.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "d65e600a08a72ca902fd34585b437e27e0d86471d12247523cf435e383c58e36", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "5f307bd2626367a1bf2ad345a3dde74bb5fd7762ed3a1a93ffe8d68fce4e7014", "generation_seconds_batch": 19.707248407998122, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.059352397918701, "target_z": 0.9921876251564564, "ce": 5.526196479797363, "genuineness": 3.911123275756836, "identity": 0.46500733494758606, "hypothesis": "La puerta sigue abierta. He dejado tu abrico sobre la silla. Dime cuándo estás listo.", "wer": 0.06666666666666667, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/d65e600a08a72ca902fd34585b437e27e0d86471d12247523cf435e383c58e36.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "d673f806025d2c862e8ccc1a14c44dfd594fbb2d8a95a04b6e19e0d4165708a4", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "10b5fc405f11d2bf0f576b004f1ef4cf45f9586e45659a79c8e7836fcefe597f", "generation_seconds_batch": 20.313484448939562, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8853760957717896, "target_z": 0.13476559915247094, "ce": 6.117925643920898, "genuineness": 4.444586753845215, "identity": 0.48727351427078247, "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_CART_reg/d673f806025d2c862e8ccc1a14c44dfd594fbb2d8a95a04b6e19e0d4165708a4.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "d6c4c186ea01be95db5cec84a3ff981f49f2b8f6ed747edafb955de9915e022e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8be8d1b2b55b20049e85fe950f1707c8ce663e9aa7ac8dd68a4f854999d46633", "generation_seconds_batch": 25.389631084050052, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.390800952911377, "target_z": 0.5039062744871328, "ce": 5.792276382446289, "genuineness": 4.118232727050781, "identity": 0.6466193199157715, "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_CART_reg/d6c4c186ea01be95db5cec84a3ff981f49f2b8f6ed747edafb955de9915e022e.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "d87bd2c0e4618cf5abb6934bd41f0f5b2a276e374c95c613491aaeb815f7ce1a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e9ec275a5d0113867d94bf6a6519f28a4d478eaa8495db96f4d9f6925585ddfc", "generation_seconds_batch": 25.28098876203876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5566015243530273, "target_z": 0.624999912934639, "ce": 5.831989288330078, "genuineness": 3.860914945602417, "identity": 0.5047562122344971, "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_CART_reg/d87bd2c0e4618cf5abb6934bd41f0f5b2a276e374c95c613491aaeb815f7ce1a.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "d88abcce27ac7403daad74eb5bd49cbf8a0d1cef5f9d4e14fa8592a58df9bac8", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9ccce4bf33bf7a9280b4d9db8965382a2c781b2486d88d9b13c888501b6827c2", "generation_seconds_batch": 21.071923605981283, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8240222930908203, "target_z": 0.8203125925069461, "ce": 6.0146708488464355, "genuineness": 3.8922953605651855, "identity": 0.319132924079895, "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_CART_reg/d88abcce27ac7403daad74eb5bd49cbf8a0d1cef5f9d4e14fa8592a58df9bac8.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "d8c875e51a842660935c6f4f0db24ee624079af29bb16db94665de850e3e2306", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "499f2e1ac1950c4e96b0bea0acbcca4870b96c2c83873e0541e73b9792fc0e5b", "generation_seconds_batch": 20.96266799303703, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.722402334213257, "target_z": 0.7460937255128672, "ce": 5.919358730316162, "genuineness": 4.243753433227539, "identity": 0.4799160361289978, "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_CART_reg/d8c875e51a842660935c6f4f0db24ee624079af29bb16db94665de850e3e2306.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "d958b0ab4766a6746ea37c7cfb4222c9a33e26070b83a4d420524d43c8242d6e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4f6be9405afb1205c67fdbc0604a908260ec0b7a1e7d12507e54e2c06a75694e", "generation_seconds_batch": 20.884523550979793, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2570905685424805, "target_z": 0.40624993470097925, "ce": 5.744620323181152, "genuineness": 3.653805732727051, "identity": 0.5185908079147339, "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_CART_reg/d958b0ab4766a6746ea37c7cfb4222c9a33e26070b83a4d420524d43c8242d6e.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "d96e6a38107429a788dc9228cbbed243547887ef2c8b65337cdf48ee19259b52", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "54f5012cb2c9eabed1dc3a3adc4c931dc719c9bac30a9daf5329f52f022d2143", "generation_seconds_batch": 22.1942776349606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5723413228988647, "target_z": -0.8242187799287178, "ce": 6.38003396987915, "genuineness": 4.162164688110352, "identity": 0.4577805697917938, "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_CART_reg/d96e6a38107429a788dc9228cbbed243547887ef2c8b65337cdf48ee19259b52.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "d9967184ea2e66124818b561159adedcebf59323fbe8672898587942456bcb24", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2570a02ef3de2fca098aae5a61754cb7cfad5f9718870bd478fa1a7eb9bb5150", "generation_seconds_batch": 15.195879185106605, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3104219436645508, "target_z": -0.28515629625347305, "ce": 6.09409761428833, "genuineness": 4.306513786315918, "identity": 0.6454449892044067, "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_CART_reg/d9967184ea2e66124818b561159adedcebf59323fbe8672898587942456bcb24.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "d9b22417bec339cfae8516377a809eedab53000efc3537a141b0e5d3fbb96b4e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5e3c231d0dad2c6ec84587376db6748fa51625f74a1802313ee71b509d3be18d", "generation_seconds_batch": 15.73440755798947, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.925336480140686, "target_z": -0.5664062309544523, "ce": 6.078212261199951, "genuineness": 4.293961524963379, "identity": 0.6086251735687256, "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_CART_reg/d9b22417bec339cfae8516377a809eedab53000efc3537a141b0e5d3fbb96b4e.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "da09195f7c98c46a67841f3aa7d6349e8580a6a40d49b1a3a2763c70cc5767b3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9b9f5664592525d59d7e7e3816d03ecfd5c4ce69d086561aa52012942e329868", "generation_seconds_batch": 21.343929087975994, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9361860752105713, "target_z": 0.17187503264951037, "ce": 5.855817794799805, "genuineness": 4.231201171875, "identity": 0.7094565033912659, "hypothesis": "J'ai trouvé la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'on 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_CART_reg/da09195f7c98c46a67841f3aa7d6349e8580a6a40d49b1a3a2763c70cc5767b3.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "da8652706d108e7eb68b4bc3dd24b3ba7802236a4cbcdd6befa7bcfead10c872", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5e801eef6962636d65e85fab0ba82a387f666202cb53c1a1828342d8897d1b26", "generation_seconds_batch": 16.240866589942016, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5698199272155762, "target_z": -0.09570311547722614, "ce": 5.974957466125488, "genuineness": 4.093128204345703, "identity": 0.20358166098594666, "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_CART_reg/da8652706d108e7eb68b4bc3dd24b3ba7802236a4cbcdd6befa7bcfead10c872.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "dad862f6dda67b3b6424b8952a075dd44c6959c39f5929d2ce3d20bc0a50b243", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8a753b4eb03865bb29efc8bb5e6a63803fb024e5e6a48ae46a1c5282bb9258b9", "generation_seconds_batch": 16.089923155959696, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6632643938064575, "target_z": -0.7578124619089046, "ce": 6.221180438995361, "genuineness": 4.532451152801514, "identity": 0.4504703879356384, "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_CART_reg/dad862f6dda67b3b6424b8952a075dd44c6959c39f5929d2ce3d20bc0a50b243.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "db147039f362b45c398128173fddb94cbf5001a18dcf737337de7ed510707483", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "80951d10092b0cae2f871281ca7fedc69324d26aee67dcac099cbd456fff2f3f", "generation_seconds_batch": 20.400047193979844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0456757545471191, "target_z": -0.4785155773861307, "ce": 6.070269584655762, "genuineness": 4.162164688110352, "identity": 0.4026890993118286, "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_CART_reg/db147039f362b45c398128173fddb94cbf5001a18dcf737337de7ed510707483.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "cd0b963ed422b75388d6cddd2734e654ba2cb567678b7c136c61d1888bed176a", "generation_seconds_batch": 15.461042138980702, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.33701133728027344, "target_z": -0.9960937255128672, "ce": 6.515059471130371, "genuineness": 4.231201171875, "identity": 0.31957849860191345, "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_CART_reg/db192910e9ae67de251fb2a990940eab881f66f627c78b50b638ce61f5a63679.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "db6c5b13e39ba94762f0e3cc50d5779969f5f0779ffdf472e31f62086b249265", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a26e9fbfb88b4aa08f94b2e5f1f0b9edbd9df659d1a009ef845082802f994e10", "generation_seconds_batch": 19.696733201039024, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8882031440734863, "target_z": 0.8671875380910954, "ce": 4.940424919128418, "genuineness": 3.7040140628814697, "identity": 0.04604315012693405, "hypothesis": "Er staat nog open. Ik heb je jas op de stoel gelekt. Zeg maar wanneer je klaar bent.", "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_CART_reg/db6c5b13e39ba94762f0e3cc50d5779969f5f0779ffdf472e31f62086b249265.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "db6dca0841e455bb3c99ba89e6e169ef8ac78d777156919f4db1579f91b5adff", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "40d94e83bc4b452f35ab9d5a85a469a37b3680393bd56bce2a0110e5797d7a93", "generation_seconds_batch": 21.339720066054724, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7530030012130737, "target_z": 0.03808595858614212, "ce": 6.197352409362793, "genuineness": 4.036643981933594, "identity": 0.4447038173675537, "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_CART_reg/db6dca0841e455bb3c99ba89e6e169ef8ac78d777156919f4db1579f91b5adff.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "db6ebafcb73d4552ee0c3db1966aa0d20b79cb80311ec1b0f796daac01a1ef19", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "89e5a1e9efb174ea58baacf17faf5caa9907db43160b8d73eaa1ecdc580464cd", "generation_seconds_batch": 19.32955087092705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4280869960784912, "target_z": -0.19921877992871784, "ce": 6.268836498260498, "genuineness": 4.331617832183838, "identity": 0.48258304595947266, "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_CART_reg/db6ebafcb73d4552ee0c3db1966aa0d20b79cb80311ec1b0f796daac01a1ef19.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "db9e683b79d5c1b0fd9ea5ac8dd2a97693c7d6d545be995b77cbc37c5cb50b59", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "fffaeef93db5b5697f5810556600c1e1d05ad5442d68c6a354c4b9f7c0a5b641", "generation_seconds_batch": 24.308702151989564, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8935513496398926, "target_z": 0.8710936384475062, "ce": 5.959072113037109, "genuineness": 4.143336772918701, "identity": 0.3072984218597412, "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_CART_reg/db9e683b79d5c1b0fd9ea5ac8dd2a97693c7d6d545be995b77cbc37c5cb50b59.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "dc13d304072ce60a5ea078de7b7b876a7e2fcd6e1736840ab3f82664e5a20ef7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4e524dc475dae0248cbc236aa8fb8aa373acd04fe66814ef65a15ed7d7221a3a", "generation_seconds_batch": 23.96424143598415, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9054327011108398, "target_z": 0.14941408494653838, "ce": 5.927301406860352, "genuineness": 4.375550270080566, "identity": 0.6458421945571899, "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_CART_reg/dc13d304072ce60a5ea078de7b7b876a7e2fcd6e1736840ab3f82664e5a20ef7.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "dc47bf95bd229b442ea6cfe2bd77b0accf824edfadbb8f350a34091ce9b12484", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1bdff738700a69818ffaef464ff8411495a27d07ebb0420acb527f5a7866c52e", "generation_seconds_batch": 20.573735035024583, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4014978408813477, "target_z": 0.5117188234613983, "ce": 5.919358730316162, "genuineness": 3.8797430992126465, "identity": 0.34232908487319946, "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_CART_reg/dc47bf95bd229b442ea6cfe2bd77b0accf824edfadbb8f350a34091ce9b12484.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "dc9f5304a6a489b16b68e8961aa2510d1a98b7f17dc5bc92f83413c51423691e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bcbe262517e17ac7bb632cf861a164899b2aced3f777fc95ded6648814a5148c", "generation_seconds_batch": 19.41208502301015, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3052263259887695, "target_z": 0.4414062309544523, "ce": 5.911416053771973, "genuineness": 4.13078498840332, "identity": 0.3674924969673157, "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_CART_reg/dc9f5304a6a489b16b68e8961aa2510d1a98b7f17dc5bc92f83413c51423691e.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "dca2ee34339acb173160ee0b7ab32eecb8ddd00d66c8c2a25971736e5eba8226", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "af91a06a9a5d95635610fa1cc5e0fe348a8af80807e979040dbf907423d81ea9", "generation_seconds_batch": 19.72106581600383, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9363389015197754, "target_z": 0.9023438343445684, "ce": 5.486483573913574, "genuineness": 4.017816066741943, "identity": -0.14504964649677277, "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_CART_reg/dca2ee34339acb173160ee0b7ab32eecb8ddd00d66c8c2a25971736e5eba8226.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "dd3970833886b0b216b5fdb36f8ae264e422a9d282c4dc8119f5bb3836b679f2", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "cb2886b18cd00b60dbf1bdfeb33fa25a36907f03016d75aca950e9c925c42171", "generation_seconds_batch": 19.87202376802452, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.834718942642212, "target_z": 0.8281249673504896, "ce": 5.935244083404541, "genuineness": 4.231201171875, "identity": 0.45425188541412354, "hypothesis": "I found the little box! Beside the old! We can look inside together!", "wer": 0.1875, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/dd3970833886b0b216b5fdb36f8ae264e422a9d282c4dc8119f5bb3836b679f2.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "df7de0dbccad769a6d31d2ae442052eadb71b904ed4fc90b781f203e8dca3b89", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "345aafca53b7d8a35b178eb0bdef4447e72fb4cfaf7cde3b322b58db52b8a42a", "generation_seconds_batch": 16.569732006057166, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0538511276245117, "target_z": 0.2578125489742656, "ce": 5.9431867599487305, "genuineness": 3.9676074981689453, "identity": 0.4433123469352722, "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_CART_reg/df7de0dbccad769a6d31d2ae442052eadb71b904ed4fc90b781f203e8dca3b89.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "dfd9c1e79af64288239871fb7f34689420df008ec7e993e76b72cb540413b4b1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "19a31142a3404c9ef3df5bb7f68b410d25d10586a86826b965c8ef09a722fc2b", "generation_seconds_batch": 19.282007831032388, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.706357002258301, "target_z": 0.7343749020514688, "ce": 5.824047088623047, "genuineness": 4.061748504638672, "identity": 0.45136719942092896, "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_CART_reg/dfd9c1e79af64288239871fb7f34689420df008ec7e993e76b72cb540413b4b1.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "e0401622c3d1106b28a21efa0ebb639468adcc1f64b601333e3641aa18e2ea61", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ffbd2e15ff21aa8ec0cba194bb5c64dbe40dc6a4c13cb1cac9a4db71ba7b8155", "generation_seconds_batch": 16.792897250968963, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3212716579437256, "target_z": 0.45312505441585066, "ce": 6.117925643920898, "genuineness": 4.080576419830322, "identity": 0.12159651517868042, "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_CART_reg/e0401622c3d1106b28a21efa0ebb639468adcc1f64b601333e3641aa18e2ea61.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "e05e6aa0cdb9322d75601f9a62d8cffad215783b1b2260e70daa8964325a8e11", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "2ee07ab28cb8e44c7a29e6d6dbaecdd3dcb4fbec52d0f4bff35316195e1c7cc3", "generation_seconds_batch": 19.90601209597662, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3799513578414917, "target_z": -0.23437498911682988, "ce": 5.935244083404541, "genuineness": 4.193545341491699, "identity": 0.583899974822998, "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_CART_reg/e05e6aa0cdb9322d75601f9a62d8cffad215783b1b2260e70daa8964325a8e11.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "e088d12c0e6e6b16a69f943f387cede24997394edb896838fece9a6345147983", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c78217d660d03a96f67c71db976e763ab572926fcdaddc3fa4b73749df32a462", "generation_seconds_batch": 25.032887223991565, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2089550495147705, "target_z": 0.3710938125782282, "ce": 5.95112943649292, "genuineness": 4.231201171875, "identity": 0.2546304166316986, "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_CART_reg/e088d12c0e6e6b16a69f943f387cede24997394edb896838fece9a6345147983.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "e18236e48e5be5dce16800fc269b5e967db79b616b8c026e156ba06db3daa9dd", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9f69b9b7788f715893072121f51e214ebc6dd561ad198f7dfec22547a1f74a69", "generation_seconds_batch": 21.610666926018894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2731359004974365, "target_z": 0.4179687581623776, "ce": 5.855817794799805, "genuineness": 3.9927117824554443, "identity": 0.0780993178486824, "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_CART_reg/e18236e48e5be5dce16800fc269b5e967db79b616b8c026e156ba06db3daa9dd.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "e1a4b8a15d27091464b70156ff07a705a00f459cf72acfb5ee003b97f7f2d87e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "1fbbb78dab39b46ec867817d215d9034e9d1f15ff1a12496b6175adf5466d128", "generation_seconds_batch": 15.40977408294566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2116291522979736, "target_z": 0.3730468627564336, "ce": 5.8875885009765625, "genuineness": 3.9173994064331055, "identity": 0.5888444185256958, "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_CART_reg/e1a4b8a15d27091464b70156ff07a705a00f459cf72acfb5ee003b97f7f2d87e.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "e261094923b2649c17600ce5dde2505a4c7224b7943bc085c2f9be2a02ab4187", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "15807aa1acedb34ca1d35e1047754b30d1dd56b8bdf1e47da2e17f21ba0153da", "generation_seconds_batch": 16.132118637906387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.061720848083496, "target_z": -0.46679692805545436, "ce": 6.165581703186035, "genuineness": 4.400654315948486, "identity": 0.6097308397293091, "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_CART_reg/e261094923b2649c17600ce5dde2505a4c7224b7943bc085c2f9be2a02ab4187.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "e2788894cdc41374fc908256666e9c37b7c82d8c0044987022de8de54382bd1e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "71f73b399e3ee137443f431f2b5a7277a4b64ade2c648ddcd9da30b1d5cebbd0", "generation_seconds_batch": 20.45752632303629, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.6849634647369385, "target_z": 0.7187499782336597, "ce": 5.863759994506836, "genuineness": 4.055472373962402, "identity": 0.6909944415092468, "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_CART_reg/e2788894cdc41374fc908256666e9c37b7c82d8c0044987022de8de54382bd1e.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "e2e6052c2a5fdeaeee574a3986ff3f98ff1aba53c57cbdf248618946e9e982d5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c76743c5db52f75b1551e489d909be6b2b2dc256a3cb56e231c54bd98be637de", "generation_seconds_batch": 15.819316926994361, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5925506353378296, "target_z": -0.07910157950495332, "ce": 6.2767791748046875, "genuineness": 4.469690799713135, "identity": 0.2892303466796875, "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_CART_reg/e2e6052c2a5fdeaeee574a3986ff3f98ff1aba53c57cbdf248618946e9e982d5.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "e351d661c351dc50ca4980fb3da314856c0b7131a9b78910943ce713b404753e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f23e91b9ef01f1e98b54091615f280635d74174c558b0fc881bec1b5d39e6ede", "generation_seconds_batch": 15.528549426002428, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0832672119140625, "target_z": 0.27929679745741287, "ce": 5.959072113037109, "genuineness": 3.8671910762786865, "identity": 0.4910849630832672, "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_CART_reg/e351d661c351dc50ca4980fb3da314856c0b7131a9b78910943ce713b404753e.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": "e3d50f3b6f073db4c3d754797446dd5260a18128f72dd18841fd327a06cabcfc", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Your message arrived after I had left.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "23aecef621169bd8644b0dedaba16d0acbd018711937b467b4321051be0893bf", "generation_seconds_batch": 20.354535614023916, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.957426905632019, "target_z": -0.5429687581623776, "ce": 6.38003396987915, "genuineness": 4.268857479095459, "identity": 0.3082992434501648, "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_CART_reg/e3d50f3b6f073db4c3d754797446dd5260a18128f72dd18841fd327a06cabcfc.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "e492391fe775f6c659946ee253f8713c45965783dc3f853dddb6202850afc6fc", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "9b3a3e087a6d38bfe332a142d35ca119d3b0e51a1e96c10848bbe7159ecd79dd", "generation_seconds_batch": 15.284291871939786, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8023229837417603, "target_z": -0.6562500217663403, "ce": 6.109982967376709, "genuineness": 4.344170093536377, "identity": 0.6664365530014038, "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_CART_reg/e492391fe775f6c659946ee253f8713c45965783dc3f853dddb6202850afc6fc.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "e4e881eddffb53d9f4d85a88ecded13b7d65fa509b7adbf959356a2bd9470b43", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "3b2cfb2c79fcac33dfcf58e56b22f66e61680132d3ee6b4ad8b578adc00356d5", "generation_seconds_batch": 19.696733201039024, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7491445541381836, "target_z": 0.7656250979485312, "ce": 5.71682071685791, "genuineness": 4.024092197418213, "identity": -0.057055793702602386, "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_CART_reg/e4e881eddffb53d9f4d85a88ecded13b7d65fa509b7adbf959356a2bd9470b43.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "e5692fc3d48e4203be6f18e07e6269374a1b91d321949d25d40223e42b13b2c7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "96b63953aa7b393806eae93c8437bc784c99531f3b79efda82367fa0d8235b0c", "generation_seconds_batch": 24.18821112893056, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.8775062561035156, "target_z": 0.8593749891168299, "ce": 5.5659098625183105, "genuineness": 3.9425034523010254, "identity": 0.4357028603553772, "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_CART_reg/e5692fc3d48e4203be6f18e07e6269374a1b91d321949d25d40223e42b13b2c7.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "e5d73cd543b3be05be9e2cd16ab920a864984d2045eb3a5b86e0e2a42a4a6ede", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "de5f9ac74a4494e56b7b44dc1c0e948631fecc3644c1eaad265dd829477fec66", "generation_seconds_batch": 19.518725777044892, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1178791522979736, "target_z": -0.42578130713664314, "ce": 6.308549880981445, "genuineness": 3.9299514293670654, "identity": 0.16161668300628662, "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_CART_reg/e5d73cd543b3be05be9e2cd16ab920a864984d2045eb3a5b86e0e2a42a4a6ede.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": "e5ff9f2d7a8f601580dac38b1e9fe60b135b3d94cd358c0d2e14e3c8567fc2dc", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Escribí el número en el sobre.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "067bf1b00b8981b0ddc3d3806d6f8f11b213b2c611c30049e1a26db03d966070", "generation_seconds_batch": 15.572665825951844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.232870101928711, "target_z": -0.34179684099009333, "ce": 6.165581703186035, "genuineness": 4.231201171875, "identity": 0.6580814123153687, "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_CART_reg/e5ff9f2d7a8f601580dac38b1e9fe60b135b3d94cd358c0d2e14e3c8567fc2dc.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "e625da92f807467cf5d8a274dbefeb62b568503cae6d915c1b8c3d1429914a1b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "762c199443a293ff9754167d6e356c86bcdd479f598458b00cf65ef0017a61e0", "generation_seconds_batch": 16.120895523927175, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.331968307495117, "target_z": 0.4609374292593942, "ce": 5.974957466125488, "genuineness": 4.20609712600708, "identity": 0.1508624404668808, "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_CART_reg/e625da92f807467cf5d8a274dbefeb62b568503cae6d915c1b8c3d1429914a1b.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "04846c1e64443e580f7be86371a70a6ee303e9fa6aec6c138ec680efec9bc90d", "generation_seconds_batch": 16.4097449160181, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5349024534225464, "target_z": -0.8515625272079254, "ce": 6.395919322967529, "genuineness": 4.419482707977295, "identity": 0.4702567458152771, "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_CART_reg/e6f1e80cc6606ef00fdc322d5829b94fd65ad80a74ab015f88321e5dc194804c.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": "e70d02b681ce6cec10b7e2a0afcf6cbedeba46fac9160806834450e9c81ea6df", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "684f10769a35d0df011bd476dbef4623a66827395fe5f58131742b1e1906605e", "generation_seconds_batch": 15.188815360073932, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7969745993614197, "target_z": -0.6601562527207925, "ce": 6.109982967376709, "genuineness": 4.319066047668457, "identity": 0.7268437147140503, "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_CART_reg/e70d02b681ce6cec10b7e2a0afcf6cbedeba46fac9160806834450e9c81ea6df.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "e7154d9ea66cbcf8a7ce08db244d7d7a47d2025b69c7708ec0366fc8099b7e8e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "abcb62ea9beab693cd00beb4950c01ee09a83f42e9a77cae0ed51657bce04060", "generation_seconds_batch": 21.810891795088537, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.770538091659546, "target_z": 0.7812500217663403, "ce": 5.816103935241699, "genuineness": 3.8044304847717285, "identity": 0.5051988363265991, "hypothesis": "The door is still open. I put your coat on the chair. Tell me when you are ready to", "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_CART_reg/e7154d9ea66cbcf8a7ce08db244d7d7a47d2025b69c7708ec0366fc8099b7e8e.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "e7a7d8e14584d7c4f8ffaf9a7d5efbffaca06c2e8471c77c97d6ab4879dc2294", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "52adf7260633f2cc032a83ba26bcde1d19d32101bbf177351d228350f5a95f5a", "generation_seconds_batch": 21.29423250700347, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.091137170791626, "target_z": -0.44531250544158507, "ce": 6.157639026641846, "genuineness": 4.005263805389404, "identity": 0.29685741662979126, "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_CART_reg/e7a7d8e14584d7c4f8ffaf9a7d5efbffaca06c2e8471c77c97d6ab4879dc2294.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "e7d3b29f2a6b8d41725440db8eecbdbae62753f48fbb5d70b22774940b5bfe2d", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ff2c773cc6cf5626b92c9b9d889ed2c6d160bb0c9ba1214e5404cef302b0cdf8", "generation_seconds_batch": 22.12807244004216, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2168247699737549, "target_z": -0.3535156644514917, "ce": 6.300607204437256, "genuineness": 4.595211505889893, "identity": 0.30925431847572327, "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_CART_reg/e7d3b29f2a6b8d41725440db8eecbdbae62753f48fbb5d70b22774940b5bfe2d.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "e7edf1dea2d977097a5a11cc5c1e6e8bbc9f1f3fd1c96b8dbc2f94db0cc6fc7b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "851d20f29655f6293cb7372473077c7e414fa97b25ab86557a86854106b3f4d5", "generation_seconds_batch": 19.882928965962492, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8800277709960938, "target_z": 0.13085941173069918, "ce": 6.229123115539551, "genuineness": 4.268857479095459, "identity": 0.6651531457901001, "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_CART_reg/e7edf1dea2d977097a5a11cc5c1e6e8bbc9f1f3fd1c96b8dbc2f94db0cc6fc7b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "e802c810a8a423a204c3f0509cf12879e5732bdcc5877a74144196aecd854238", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "416b119379da173fed8d8871831ab78c3f4a83374b028749d868fd7a1aebe2a5", "generation_seconds_batch": 19.29071022104472, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7743966579437256, "target_z": 0.053710969469312246, "ce": 6.3641486167907715, "genuineness": 4.105680465698242, "identity": 0.3600701689720154, "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_CART_reg/e802c810a8a423a204c3f0509cf12879e5732bdcc5877a74144196aecd854238.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "e842fcd91ee7d3a4572afec0eb29798adc17b915bd9ddd963f7ae1c1bb4b6607", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "231cea2ca84a29982cbfa16d5e52497d33d8aea07768adc78229f412cd08bf7b", "generation_seconds_batch": 22.12807244004216, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.337164044380188, "target_z": -0.26562501088317014, "ce": 6.300607204437256, "genuineness": 4.482243061065674, "identity": 0.2446533888578415, "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_CART_reg/e842fcd91ee7d3a4572afec0eb29798adc17b915bd9ddd963f7ae1c1bb4b6607.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "e850a3063f827a5fc7505af096eff225d174a1c4751dd1ee3b15faba4790c278", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cbcdc651a1d9b1d9ceaf36229d88c2e6fa4025d743d9ad169faee5cc48df5940", "generation_seconds_batch": 20.78243125299923, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5883865356445312, "target_z": -0.8125000435326805, "ce": 6.403861999511719, "genuineness": 4.13078498840332, "identity": 0.3981912434101105, "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_CART_reg/e850a3063f827a5fc7505af096eff225d174a1c4751dd1ee3b15faba4790c278.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "e8525c3b16b76ebb48bee98adaa351b5cb2eeb139976e9367aa5935956c0127b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f5d23be4068870ca27a8573db3b111ebf8752f10aa4623c64971e9ea99995c7b", "generation_seconds_batch": 18.965767242014408, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.045828342437744, "target_z": 0.2519530501782054, "ce": 6.078212261199951, "genuineness": 4.532451152801514, "identity": 0.5104526281356812, "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_CART_reg/e8525c3b16b76ebb48bee98adaa351b5cb2eeb139976e9367aa5935956c0127b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "e8ecb83dc6aba6fcfc823ee9020a3e3ff2ec79f5c8bab71fd4a48bbaf6200126", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7b37a70f51f85b248d6a7ab7645a47ebb617da9d092b56e7331c6a215daf3c65", "generation_seconds_batch": 18.98300054098945, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7356206178665161, "target_z": 0.025390610035641078, "ce": 6.086154937744141, "genuineness": 4.143336772918701, "identity": 0.41328513622283936, "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_CART_reg/e8ecb83dc6aba6fcfc823ee9020a3e3ff2ec79f5c8bab71fd4a48bbaf6200126.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "e931c32cc04a8a56ea7ed8cde9fc1b3029f9300ebd41da1949006fc15be861de", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5422e8d6516f2b4b03591ce50a007570227fabc7a593f54cd7ab71025d9cc997", "generation_seconds_batch": 20.382365124998614, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.786583423614502, "target_z": 0.7929688452277386, "ce": 5.5817952156066895, "genuineness": 3.8044304847717285, "identity": 0.40872448682785034, "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_CART_reg/e931c32cc04a8a56ea7ed8cde9fc1b3029f9300ebd41da1949006fc15be861de.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "e9b744a217527495f3e58bc798047756cd3117fa87f300871d059503f2adfff7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "893bca2cfbe1309e3745a77c6537458c921303d55b8c70a9fa965279a5e7ee78", "generation_seconds_batch": 15.700251981033944, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0379586219787598, "target_z": 0.9765625272079254, "ce": 5.609594821929932, "genuineness": 4.13078498840332, "identity": 0.2947580814361572, "hypothesis": "¡He encontrado la cajita! Estaba junto a la ventana fija. 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_CART_reg/e9b744a217527495f3e58bc798047756cd3117fa87f300871d059503f2adfff7.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "e9be662749a372ce25bbc273bd0111deb0aead2ca6f34ea675a8b35f1740650f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7ea5af4300f41fbd38c325f558c9f4ec7687bb6d2b9d46555dd30bb73aa82614", "generation_seconds_batch": 22.981363175902516, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9722877740859985, "target_z": 0.19824216777425416, "ce": 6.1894097328186035, "genuineness": 4.068024158477783, "identity": 0.40416860580444336, "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_CART_reg/e9be662749a372ce25bbc273bd0111deb0aead2ca6f34ea675a8b35f1740650f.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "4cc857215610273b303728b3d90e25618de81e79dfa45d34509d9a6044d7ef41", "generation_seconds_batch": 16.59513765689917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7274452447891235, "target_z": -0.7109375163247552, "ce": 6.348263263702393, "genuineness": 4.2751336097717285, "identity": 0.31724703311920166, "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_CART_reg/e9d266c3907611a53d0c5f7688a7e231bba29ead6f87b8ae8d29cb37ab2fae41.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "ea0b4b969a7412610627d9b8452ad453d589ceb7b87b8c736a07346c2b619655", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fb2374693d4b8bfb35d4fbcfef1833eba82f5dbd674d741cd6d22a65ce1960bf", "generation_seconds_batch": 19.317320410045795, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4976164102554321, "target_z": -0.14843747279207467, "ce": 6.24500846862793, "genuineness": 4.18099308013916, "identity": 0.4468076825141907, "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_CART_reg/ea0b4b969a7412610627d9b8452ad453d589ceb7b87b8c736a07346c2b619655.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "ea0e74f7879cfdab2b4697cd18f4f189204845e080bf70268f3458b9848128d0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "61af932a715ce530d36b272cb5ae5497b1eac68b9003fa5fac2d45069728a2b1", "generation_seconds_batch": 15.022237317985855, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2570905685424805, "target_z": 0.40624993470097925, "ce": 5.903473377227783, "genuineness": 4.061748504638672, "identity": 0.5674989819526672, "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_CART_reg/ea0e74f7879cfdab2b4697cd18f4f189204845e080bf70268f3458b9848128d0.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "eaa41e3eda76be47dfe43fda68e746b50da0a27825f155f196663840c60d49f3", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "08ee1c03fefe03a8467cf07bbee4f6980ef57fc072372bca017f67db17cf67f5", "generation_seconds_batch": 20.769450145075098, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.219651699066162, "target_z": 0.3789061874217718, "ce": 6.284721851348877, "genuineness": 3.8169827461242676, "identity": 0.15348675847053528, "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_CART_reg/eaa41e3eda76be47dfe43fda68e746b50da0a27825f155f196663840c60d49f3.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": "ead51552d374caac90e9ed98eeb02f246725c0638f19c690ff82f5727b05d2ae", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ed01f7678666fc08a38f3708346859580eb2d320e595882f59b070cbf9bf4e17", "generation_seconds_batch": 20.448546694009565, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8290650844573975, "target_z": -0.6367187363960374, "ce": 6.260893821716309, "genuineness": 4.582659721374512, "identity": 0.4941956400871277, "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_CART_reg/ead51552d374caac90e9ed98eeb02f246725c0638f19c690ff82f5727b05d2ae.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": "ebe61f97499d406896f3bc92951d41a5f07dd5562ae8027067a5ea5a22c59972", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7cd84c51b020188c18e1a3612222cd6fe96e12dbca5bf9b89f2936dbdea38d00", "generation_seconds_batch": 20.417473209905438, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9948657751083374, "target_z": -0.5156250108831701, "ce": 6.348263263702393, "genuineness": 4.331617832183838, "identity": 0.4982612133026123, "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_CART_reg/ebe61f97499d406896f3bc92951d41a5f07dd5562ae8027067a5ea5a22c59972.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "b4a1f0eee773b0f0af45dd1061d65736928c80951b59d2e18b3631f51247b2ed", "generation_seconds_batch": 15.92157692706678, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.21399784088134766, "target_z": -1.085937516324755, "ce": 6.149696350097656, "genuineness": 4.394378662109375, "identity": 0.8773958683013916, "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_CART_reg/ebe89981a89e9506d02bacdfa583f9bfe1f04faa699193749ab6419de85c588f.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "eea490d0549043468379b7610ef9ceb2122fcc86c14b9a7dc5c6064641cb2728", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4c85fa26268dc08ff8f567b466956ad3d0fdd2e65ad67616137357a5f875e0f7", "generation_seconds_batch": 20.91190315701533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3478608131408691, "target_z": -0.2578125489742656, "ce": 6.268836498260498, "genuineness": 4.444586753845215, "identity": 0.6438695192337036, "hypothesis": "J'ai trouvé la petite botte. Elle était près de 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_CART_reg/eea490d0549043468379b7610ef9ceb2122fcc86c14b9a7dc5c6064641cb2728.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "eeba5506d5f561b72f749311974e7105e6a7d535189430183aa090bd3ef75e38", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5b2922ec776a07bbff6bf970137d7f5414bed0eab9dffcf98bbc16dfab4fdfce", "generation_seconds_batch": 15.030100949923508, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.471027135848999, "target_z": 0.5625000435326805, "ce": 5.561938762664795, "genuineness": 3.8546390533447266, "identity": 0.2172582894563675, "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_CART_reg/eeba5506d5f561b72f749311974e7105e6a7d535189430183aa090bd3ef75e38.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "ef0b3bf703606cb16bcad81cc4006dccbbafab3f2cf5ff22f8a5b5b09fc4d668", "bucket": "vn_S_CART_reg", "phase": "wording", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "cb8de0f9f9bdbbd6f6ac3246544b7a46fff4c3dcfd66367b120965e058d37b73", "generation_seconds_batch": 17.28488218097482, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3852996826171875, "target_z": -0.2304688016950581, "ce": 6.2767791748046875, "genuineness": 4.406930446624756, "identity": 0.5575779676437378, "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_CART_reg/ef0b3bf703606cb16bcad81cc4006dccbbafab3f2cf5ff22f8a5b5b09fc4d668.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "ef339309f37a08b61976160a2ec03954f2c020cf1d852988d5d08f24af87640b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fee5479850e468181662e838d09880be2823a17b860098ebf5b186d3b2945223", "generation_seconds_batch": 14.943225278984755, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2676347494125366, "target_z": -0.3164062309544523, "ce": 6.332377910614014, "genuineness": 4.444586753845215, "identity": 0.6419013738632202, "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_CART_reg/ef339309f37a08b61976160a2ec03954f2c020cf1d852988d5d08f24af87640b.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "ef72af5fa0c66ff679f6bc3201e161245ba57495f891cf7039ea213f6f65b731", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "d40e41e687ec5e35e66a47bf59dffc3ebefa2109c771e3d94e52bcaf40f6fca6", "generation_seconds_batch": 19.832552331034094, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9092912077903748, "target_z": -0.5781250108831701, "ce": 6.197352409362793, "genuineness": 4.319066047668457, "identity": 0.35513725876808167, "hypothesis": "The door is still open. 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_CART_reg/ef72af5fa0c66ff679f6bc3201e161245ba57495f891cf7039ea213f6f65b731.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": "ef87c3583ec42b18f8482668e16eab3c827aac7ee1198f5434cd2c76b3e24990", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Escribí el número en el sobre.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "08de274ed08ee5180c49328702d42ec4477d3fdafa9bb6f3b117fc3d87dc2511", "generation_seconds_batch": 20.051023079897277, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9627753496170044, "target_z": -0.5390624836752448, "ce": 6.181467056274414, "genuineness": 4.218649387359619, "identity": 0.5154456496238708, "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_CART_reg/ef87c3583ec42b18f8482668e16eab3c827aac7ee1198f5434cd2c76b3e24990.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "ef9cdabfbcbef97d39f0e21d337b3365ec870b5d694dbb9c9de53437ce514d73", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "f6ab69b19c717fbc517018a8c9613ff12457238c08c3847e970c8eb7f0128efd", "generation_seconds_batch": 26.60732024302706, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5245110988616943, "target_z": 0.6015624401425643, "ce": 5.704906940460205, "genuineness": 3.8420867919921875, "identity": 0.1640508770942688, "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_CART_reg/ef9cdabfbcbef97d39f0e21d337b3365ec870b5d694dbb9c9de53437ce514d73.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "efd78cd957e7b0d045b9bb65d4038774301d55ea62cc938598e019df1ffe0d33", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "7fa92ddc0d73db487f9da557740ce9a8d7e73790efca569bff6905e47d412dc5", "generation_seconds_batch": 20.95564452605322, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.765036940574646, "target_z": 0.04687503264951038, "ce": 6.23706579208374, "genuineness": 4.582659721374512, "identity": 0.7367344498634338, "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_CART_reg/efd78cd957e7b0d045b9bb65d4038774301d55ea62cc938598e019df1ffe0d33.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "eff8cf618db6015927db753637289784e2fccc4f96e70b189f9f7d490528b016", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c7b3db8f287a2bdccfc76e21c6c19eb2ff4f88fa5bfe8aaf21810873eac5cbd9", "generation_seconds_batch": 23.823033366934396, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5138144493103027, "target_z": 0.5937500652990207, "ce": 5.8875885009765625, "genuineness": 4.049196243286133, "identity": 0.7621163725852966, "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_CART_reg/eff8cf618db6015927db753637289784e2fccc4f96e70b189f9f7d490528b016.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "f193033f2d4af147aa0240a0806348537de3c62f9ee35e8199bae96d65d8dd92", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2171651b89e4b9dc5e3f6c8d27f2706e53f835585c25e9bc5bac3e166aec1cd6", "generation_seconds_batch": 19.199448053957894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0538511276245117, "target_z": 0.2578125489742656, "ce": 6.260893821716309, "genuineness": 4.268857479095459, "identity": 0.34878677129745483, "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_CART_reg/f193033f2d4af147aa0240a0806348537de3c62f9ee35e8199bae96d65d8dd92.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "4bcaacbc2beca469dd4244afaf635a49f8706e7ac8cf529c15506ccac4340b55", "generation_seconds_batch": 16.205504458979703, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8183682560920715, "target_z": -0.6445312418376224, "ce": 6.284721851348877, "genuineness": 4.406930446624756, "identity": 0.4780331254005432, "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_CART_reg/f21bc0adf2df6aeda4aa5bc68657c60b57559ffed3b647118e970ebfd0ba6ad8.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "f2ae13711298e8ef2ecb3d3a974cdb6e10bba337c950d23b119c4469fdcf609c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "6f12bd848a19fe6166582144a41da76f97dac912ed78ea97c7f57a0d03576b03", "generation_seconds_batch": 21.0297713939799, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.0165648460388184, "target_z": 0.9609374292593942, "ce": 5.80021858215332, "genuineness": 3.7918784618377686, "identity": 0.07711824774742126, "hypothesis": "Ik heb het kleine doosje en van de huh. Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.2222222222222222, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/f2ae13711298e8ef2ecb3d3a974cdb6e10bba337c950d23b119c4469fdcf609c.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "f2faed4d69bc67ec48775403c57b41213c9698e8cb8fee354c3da0cdb9793a78", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "cdb3301b9638d07815bfa453311c4f1146f9d4c9881daf5b6b715f0a09eae265", "generation_seconds_batch": 20.200493552954867, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.024434804916382, "target_z": 0.23632812636039627, "ce": 5.9431867599487305, "genuineness": 3.986435890197754, "identity": 0.38413316011428833, "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_CART_reg/f2faed4d69bc67ec48775403c57b41213c9698e8cb8fee354c3da0cdb9793a78.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "f37d928bab48f78ed93505a58d4fd150f3b724d9ccc39851df3f26181a414e34", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "cb1c2c8db4d7cf896af8ade5322248bb68c59d889d4ce10829ce9ca3a1fb62da", "generation_seconds_batch": 17.266092004952952, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9737777709960938, "target_z": 0.929687581623776, "ce": 5.331601142883301, "genuineness": 3.7793264389038086, "identity": 0.43223586678504944, "hypothesis": "The door's still open! I put your coat on the chair! Tell me when you are ready!", "wer": 0.1111111111111111, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/f37d928bab48f78ed93505a58d4fd150f3b724d9ccc39851df3f26181a414e34.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "f47843affa2f36d3a8e166dde7826aa33a1c3f6f59140bc6b056aecb987ea79a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "fa1f9ab0c43a3f21b8a53a0b8b0a0d70e54aa21c71f904c39e9ebdf700192688", "generation_seconds_batch": 21.848669525934383, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6914962530136108, "target_z": -0.006835936819801867, "ce": 6.260893821716309, "genuineness": 4.256305694580078, "identity": 0.5699453353881836, "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_CART_reg/f47843affa2f36d3a8e166dde7826aa33a1c3f6f59140bc6b056aecb987ea79a.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "f557cf38b7e6a4c9396875cf53c4b53f7c7cc1946fd6bab200edf1b186ac9a7b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "18173555823f3d5d329b74c6428b42b7d5f52516cd3dfeab4df6bbd12560b93b", "generation_seconds_batch": 21.29423250700347, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.489593744277954, "target_z": -0.15429688452277385, "ce": 6.086154937744141, "genuineness": 3.7040140628814697, "identity": 0.2453698068857193, "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_CART_reg/f557cf38b7e6a4c9396875cf53c4b53f7c7cc1946fd6bab200edf1b186ac9a7b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "f570e0f44dfc0d411daaf05204c72eeb8cfa673920d159cf3b9f58c37c5cf6ef", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "225075cfd8725bdbdb43038a69312c78aa810fc90b60495fd0ad9ae600bfc226", "generation_seconds_batch": 14.795922335004434, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7329463958740234, "target_z": 0.023437472792074684, "ce": 6.149696350097656, "genuineness": 4.281409740447998, "identity": 0.3967306315898895, "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_CART_reg/f570e0f44dfc0d411daaf05204c72eeb8cfa673920d159cf3b9f58c37c5cf6ef.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "f58b99747db2077cb4c0ace452d147d0499b26e27ea8cd227585eec25a240fae", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c145d6b3f861399e8a56a833d8e2f76e5141ee86779c619785db7f0d5aff8b01", "generation_seconds_batch": 21.83409894502256, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6005733013153076, "target_z": -0.07324216777425414, "ce": 6.125868320465088, "genuineness": 4.406930446624756, "identity": 0.6278978586196899, "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_CART_reg/f58b99747db2077cb4c0ace452d147d0499b26e27ea8cd227585eec25a240fae.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "f5de32a165d6cba4901058412f0fbc8099f528662c697a557d564c3dc88a26d5", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d953f9cbd9ca4c20be50e0f38b6cb0ed51662b6087cdfca9ed749107e15f73e1", "generation_seconds_batch": 19.8963846139377, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.503117561340332, "target_z": 0.5859375163247552, "ce": 5.831989288330078, "genuineness": 4.137060642242432, "identity": 0.1666146218776703, "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_CART_reg/f5de32a165d6cba4901058412f0fbc8099f528662c697a557d564c3dc88a26d5.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "f5f608f41685b992daeb9d474cde710fb0139461a368f56071656e02c64d569a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c747bde02b75d1cf932eee02985593ba71255bb7e152e64a49fec9615d9365b1", "generation_seconds_batch": 22.299648517975584, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.230348587036133, "target_z": 0.3867187363960373, "ce": 6.0305562019348145, "genuineness": 4.162164688110352, "identity": 0.5108757019042969, "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_CART_reg/f5f608f41685b992daeb9d474cde710fb0139461a368f56071656e02c64d569a.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "f605caa20bc7f856a50799d5a1de7a95f9b9cd6f125b452d071d08a7b27ef05c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5f3472e9bdfe424306c35cf32a56afa6e4fdf3c5785edb1d7348e61834313b1f", "generation_seconds_batch": 21.09462536103092, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6139442920684814, "target_z": -0.06347656862178319, "ce": 6.125868320465088, "genuineness": 4.419482707977295, "identity": 0.6262983083724976, "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_CART_reg/f605caa20bc7f856a50799d5a1de7a95f9b9cd6f125b452d071d08a7b27ef05c.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "f6217df6e474f763a71d5cb24d49cac377fbf6e9e66a5bdd092c2ffd8baf3c03", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b70cb18be10629ac9a87b6f8a4ee45a168d951bcf9adae4d90b9e6296d9faf65", "generation_seconds_batch": 19.8977074889699, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4427951574325562, "target_z": -0.1884765686217832, "ce": 6.2767791748046875, "genuineness": 4.193545341491699, "identity": 0.29645493626594543, "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_CART_reg/f6217df6e474f763a71d5cb24d49cac377fbf6e9e66a5bdd092c2ffd8baf3c03.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "f625f6337e45b1d6fe9a998cb2d493a6c4a8dc7fe63e74d7e0d62085626789cd", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0c04698dd55ac97c7dca9b97635889f2ac2e16abf2d848fc7df5e08d837b05be", "generation_seconds_batch": 16.040086375083774, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.962928056716919, "target_z": 0.19140623095445228, "ce": 6.157639026641846, "genuineness": 4.337893962860107, "identity": 0.3149663805961609, "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_CART_reg/f625f6337e45b1d6fe9a998cb2d493a6c4a8dc7fe63e74d7e0d62085626789cd.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "f66141d8d274e7fad776585a431211008e7b42d69033356d72c302604f02e019", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "8a0f3fc5a600ea9581684878feb47e1f3fb584838846638de0a3428a3fa1c157", "generation_seconds_batch": 19.734027337050065, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 3.1128363609313965, "target_z": 1.0312500217663403, "ce": 5.911416053771973, "genuineness": 4.086852550506592, "identity": 0.5320895314216614, "hypothesis": "Bien que he encontrado la cajita. Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.11764705882352941, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/f66141d8d274e7fad776585a431211008e7b42d69033356d72c302604f02e019.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "f6d425d248085d544c35f123c80a427843c1c96d785c1ec46238e5421b4b4c24", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5cb1f6fa83df57351bdfcbfc03e9081e579383d62a039a9b8d8205b341da3a7e", "generation_seconds_batch": 20.549985943944193, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4386310577392578, "target_z": -0.9218750326495104, "ce": 6.260893821716309, "genuineness": 4.080576419830322, "identity": 0.6422867178916931, "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_CART_reg/f6d425d248085d544c35f123c80a427843c1c96d785c1ec46238e5421b4b4c24.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": "f6dacf23e1007c9486eb39034e46192f159d01577d5143910733b1063b28998a", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "94e201b98d5b59a8b1d4c86b77a33ac2f4e80dd975c679bead155770e4272e63", "generation_seconds_batch": 20.855196572025307, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0323046445846558, "target_z": -0.4882812636039627, "ce": 6.117925643920898, "genuineness": 4.362998008728027, "identity": 0.5696936845779419, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, l'île des biens, et vérifie 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_CART_reg/f6dacf23e1007c9486eb39034e46192f159d01577d5143910733b1063b28998a.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "f6e84d1179ad93b3c9f94295e4add40d52f15e994aa7e841950b330cf9f68817", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "b34513c446c7704b978d6511742e78688acb3054d2cee54de6a0da1bc691c798", "generation_seconds_batch": 19.263538900995627, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3237930536270142, "target_z": -0.2753906100356411, "ce": 6.157639026641846, "genuineness": 4.118232727050781, "identity": 0.44581496715545654, "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_CART_reg/f6e84d1179ad93b3c9f94295e4add40d52f15e994aa7e841950b330cf9f68817.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "f7073385cdb01497f516467bd198b0c0d5044a3d61bfcd56dbaab513bd832f8b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "19928fd08bbe76135f4914b31ced253af3123c60d2d72fa3fc2337d8907b83bb", "generation_seconds_batch": 23.96424143598415, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.502964735031128, "target_z": -0.1445312853703029, "ce": 5.959072113037109, "genuineness": 4.457139015197754, "identity": 0.491495817899704, "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_CART_reg/f7073385cdb01497f516467bd198b0c0d5044a3d61bfcd56dbaab513bd832f8b.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "f721aebe9a75dcb0a0d23bef6bba3172396ac40aebfae5febdcf31061467eb27", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "436ab77ab8e55410337b2c77a384e102060fc26bbb0476b434052d23e04d4584", "generation_seconds_batch": 19.724742676015012, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4924206733703613, "target_z": 0.5781249673504896, "ce": 5.816103935241699, "genuineness": 4.21237325668335, "identity": 0.14278046786785126, "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": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/f721aebe9a75dcb0a0d23bef6bba3172396ac40aebfae5febdcf31061467eb27.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "f72a12ad9b1b1580134fde986295f20b97282bb65ae3512d21a173698010ec0b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "092050e9dd9fd86c4ad72f839ed59b1e14eeba0c64e4052b6236fad4b708ca5c", "generation_seconds_batch": 19.263538900995627, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3478608131408691, "target_z": -0.2578125489742656, "ce": 6.292664527893066, "genuineness": 4.319066047668457, "identity": 0.40843671560287476, "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_CART_reg/f72a12ad9b1b1580134fde986295f20b97282bb65ae3512d21a173698010ec0b.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": "f75a287602b29390c35fb6d70eb3ca0aa907e3bcb69302ee5aac36d97bb02a7b", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "259a080295860c6073a30ff8ee9a512e9d9be5ec733204eb118c48ce962528b1", "generation_seconds_batch": 15.302610380924307, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4695372581481934, "target_z": -0.16894528325148028, "ce": 6.332377910614014, "genuineness": 4.293961524963379, "identity": 0.7000735998153687, "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_CART_reg/f75a287602b29390c35fb6d70eb3ca0aa907e3bcb69302ee5aac36d97bb02a7b.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "f75aa40b8bbecb67fc90270ec60d726ab11653c6a321ec129602c67099e516b9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "caf142d04812d1fb1d8b3bafe61cde2720633bae650262df25503088e5f005f0", "generation_seconds_batch": 15.40977408294566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.750328779220581, "target_z": 0.03613282134257573, "ce": 6.372091293334961, "genuineness": 4.457139015197754, "identity": 0.5497704744338989, "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_CART_reg/f75aa40b8bbecb67fc90270ec60d726ab11653c6a321ec129602c67099e516b9.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "f778421fa619a77f99628b6549bfb4c812c43246769bdaedb585f29fed28535c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "856b1fb1317c63808aa3b437e1e6a14a916f98bd174a356b1781128c81b438cb", "generation_seconds_batch": 18.955261257011443, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3184446096420288, "target_z": -0.27929688452277385, "ce": 5.998785495758057, "genuineness": 4.011539936065674, "identity": 0.45352989435195923, "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_CART_reg/f778421fa619a77f99628b6549bfb4c812c43246769bdaedb585f29fed28535c.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": "f778dd9abe068858d91cab92aac115ebf8900d712390a58b54db63700c10083e", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "45d443cc1cbfac08bc7c0c0ba395d4aca41e5a4782557a4bb60e7f988320d677", "generation_seconds_batch": 20.212558539002202, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5776897668838501, "target_z": -0.8203125054415851, "ce": 6.403861999511719, "genuineness": 4.256305694580078, "identity": 0.6721158027648926, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zum kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/f778dd9abe068858d91cab92aac115ebf8900d712390a58b54db63700c10083e.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "9a178697e996f396386695e665d3975bc603d961982c578b47207a4ff369ecea", "generation_seconds_batch": 16.92112905299291, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.051024079322815, "target_z": -0.4746093899643589, "ce": 6.149696350097656, "genuineness": 4.494794845581055, "identity": 0.5260385274887085, "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_CART_reg/f7a65885d0e107603bdce9d92b5fc3b68a68524db7b0b06f79b6538794e8bf5c.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "f836a02182f9fa910a935bc051d5503f7eb1757946e44c18eb7a8ce1baabeac7", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4d0eadb6eb63916f7579d4cd06810e9ca0df0fd287a9cfb100172c5960b419d0", "generation_seconds_batch": 21.52729752101004, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6540573835372925, "target_z": -0.03417968409900934, "ce": 6.046441555023193, "genuineness": 4.3504462242126465, "identity": 0.5672764778137207, "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_CART_reg/f836a02182f9fa910a935bc051d5503f7eb1757946e44c18eb7a8ce1baabeac7.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_CART_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": "b9331cf44722dcd51ff1bdffa2dcd6df632602e3961961f075c447ab413d3212", "generation_seconds_batch": 15.132483097026125, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.11772644519805908, "target_z": -1.1562500217663403, "ce": 6.181467056274414, "genuineness": 4.457139015197754, "identity": 0.5257972478866577, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis zum kleinen Holztor. Ich warte auf der anderen Seite.", "wer": 0.09523809523809523, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/f86feddc7af334e5c3d6691e715ece3168f8e7618bd3b63ca7d0171f412fdf9d.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "f96d1eaaf0e3c78e9ee6523d5915ba7f3d5306b96826045e2d7ed6f929c1e9bf", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "336ea80b61958af75a4fb59098165701f46470a95fd1bb105890e8adfe19be0e", "generation_seconds_batch": 15.819316926994361, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3264672756195068, "target_z": -0.2734374727920747, "ce": 6.332377910614014, "genuineness": 4.482243061065674, "identity": 0.48673373460769653, "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_CART_reg/f96d1eaaf0e3c78e9ee6523d5915ba7f3d5306b96826045e2d7ed6f929c1e9bf.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": "f99b59d3f38015b672088af6bd81b13c1b4f048163ab783ec868101e85af87a1", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "30e60aabfc746eb95eb4346e289bf4d5e29975b2a53f62f132351d3d64f3bedb", "generation_seconds_batch": 15.076472377055325, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4468064308166504, "target_z": -0.1855469062891141, "ce": 6.062326908111572, "genuineness": 4.093128204345703, "identity": 0.6248302459716797, "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_CART_reg/f99b59d3f38015b672088af6bd81b13c1b4f048163ab783ec868101e85af87a1.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-0", "key": "f9baa7ea61b233e6a5d72816b2342f6e7141a9ad32f00aa5fe1c20d36ccdad88", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La porte est encore ouverte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "80ed75c55ff58cc819b4db1fe15b7eb329aec5182988e00f5e979a2fcc5807c0", "generation_seconds_batch": 20.658666242961772, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.313096284866333, "target_z": -0.28320307194454564, "ce": 6.300607204437256, "genuineness": 4.337893962860107, "identity": 0.5823269486427307, "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_CART_reg/f9baa7ea61b233e6a5d72816b2342f6e7141a9ad32f00aa5fe1c20d36ccdad88.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "f9eb9736dc9b95daea01446b66309ef15f014a1712d6ff082f429e516b0e3ef9", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "90240818e3c92c1035da3f427fb27bbb16fa3b670e0d3121750a4998a1cb6c66", "generation_seconds_batch": 20.64634263899643, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3959965705871582, "target_z": -0.22265625272079254, "ce": 6.1894097328186035, "genuineness": 4.3127899169921875, "identity": 0.47590896487236023, "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_CART_reg/f9eb9736dc9b95daea01446b66309ef15f014a1712d6ff082f429e516b0e3ef9.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "fa00a059a78dd43b03cd66ae676e4f6212178492a0ced78bda7f91c219844cf1", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "5d40f9cbb3c88f96e3b50c59ed25d33403122b27ebb50ed18bec0c8321807b7e", "generation_seconds_batch": 16.24323416291736, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9535683393478394, "target_z": 0.18457029413465043, "ce": 6.078212261199951, "genuineness": 4.413206577301025, "identity": 0.17781849205493927, "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_CART_reg/fa00a059a78dd43b03cd66ae676e4f6212178492a0ced78bda7f91c219844cf1.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "fa74cb694dde5e4f76ab052dc833eb7d78232d1249c45a32543794529b7e821b", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "62da229860efd742597b184d788b31c526abb9fa75b2405b08943d172e52cdb8", "generation_seconds_batch": 19.27111886208877, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5724941492080688, "target_z": -0.09374997823365974, "ce": 6.125868320465088, "genuineness": 4.432034492492676, "identity": 0.577358603477478, "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_CART_reg/fa74cb694dde5e4f76ab052dc833eb7d78232d1249c45a32543794529b7e821b.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": "fad3964b3a0a5de4f99e63d45549506cf191f9de81536e37381af2e232456c96", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a5c2025826bca266880066968f5817f61932f2b8329ab85a33b76e029dd8697e", "generation_seconds_batch": 20.08546127693262, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1419470310211182, "target_z": -0.40820315900990667, "ce": 6.0305562019348145, "genuineness": 4.413206577301025, "identity": 0.5493372678756714, "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_CART_reg/fad3964b3a0a5de4f99e63d45549506cf191f9de81536e37381af2e232456c96.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "fb37e27b45affc7171288d2a49eb90e04960b72f117ba5a70ee578c9c1bb6489", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "ff038d4e7dca191eaa05831056b9ffeb61a26dd575beab98b5b9f2c9bc2e4ade", "generation_seconds_batch": 14.796144105028361, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.917466640472412, "target_z": 0.15820315900990664, "ce": 6.078212261199951, "genuineness": 4.469690799713135, "identity": 0.2707083225250244, "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_CART_reg/fb37e27b45affc7171288d2a49eb90e04960b72f117ba5a70ee578c9c1bb6489.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "fb4fe49e198fa337af55c926e254a792cffb98a32c69252182ebf31aa0593f34", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "00e30091a4e774c7c3f2016d1b041a65bd904a5e5efa89602dcdff9b2f7397b0", "generation_seconds_batch": 19.982762280968018, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.454981803894043, "target_z": 0.5507812200712822, "ce": 5.49045467376709, "genuineness": 4.20609712600708, "identity": -0.020545359700918198, "hypothesis": "Ik heb een klein doosje gevonden. Het stond naast het oude raam. We kunnen samen in kijken.", "wer": 0.16666666666666666, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/fb4fe49e198fa337af55c926e254a792cffb98a32c69252182ebf31aa0593f34.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-1", "key": "fbb45b4742e6085d58c5e916cf79c285aad34f9ae6edf9716a1d351b06d4de75", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "c22e5ddeb616dd76e94082e9817cb4afff788b07149c7f1627c9dc7e077e1642", "generation_seconds_batch": 22.981363175902516, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3558834791183472, "target_z": -0.2519531372435664, "ce": 6.2767791748046875, "genuineness": 4.293961524963379, "identity": 0.182821586728096, "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_CART_reg/fbb45b4742e6085d58c5e916cf79c285aad34f9ae6edf9716a1d351b06d4de75.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "fbd0d8c178bf0caab6b79a52e21898408e20a4db500ff13499da7b4a0881ba1f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "0bca762444dfaeecc729fd1862f68f466aac01c60e486a0b61e6d09e7fb3ea2a", "generation_seconds_batch": 20.303611007984728, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8679938316345215, "target_z": 0.12207033766733091, "ce": 6.348263263702393, "genuineness": 4.055472373962402, "identity": 0.37312695384025574, "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_CART_reg/fbd0d8c178bf0caab6b79a52e21898408e20a4db500ff13499da7b4a0881ba1f.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "fc31ef33276a6760ec36dc95a64c5bb07c7d8d129c38cf3a4307b1235c6cfb8f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "db3a495de8930a739f601c64d5aa0c2e2cf317f9b272c3c84ae1dc9933e33548", "generation_seconds_batch": 19.844987511984073, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7596884965896606, "target_z": 0.04296875816237759, "ce": 6.109982967376709, "genuineness": 4.055472373962402, "identity": 0.5902482271194458, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude raam. We kunnen het 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_CART_reg/fc31ef33276a6760ec36dc95a64c5bb07c7d8d129c38cf3a4307b1235c6cfb8f.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": "fcc13e2044c892cc2accf7e99fb1cbdf390e36541815b42f58b49dbb0f495981", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8181072864426be24f74b2f52db6fbc97c3d12ca329ee07d316508da1e7a0e00", "generation_seconds_batch": 15.236427792930044, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8011386394500732, "target_z": 0.07324216777425414, "ce": 5.959072113037109, "genuineness": 4.030367851257324, "identity": 0.07730453461408615, "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_CART_reg/fcc13e2044c892cc2accf7e99fb1cbdf390e36541815b42f58b49dbb0f495981.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-0", "key": "fd503edc24e3272c478d34ee9e5960a384f78f3af83539621fd56700dee025e0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] The door is still open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] I put your coat on the chair.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "92ef7537523e747c7e86faf3c7b88a49c482dae01a5d05fff983ee4a3f50d12c", "generation_seconds_batch": 22.62403995904606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3079004287719727, "target_z": 0.44335928113265766, "ce": 5.9431867599487305, "genuineness": 4.093128204345703, "identity": 0.4350356459617615, "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_CART_reg/fd503edc24e3272c478d34ee9e5960a384f78f3af83539621fd56700dee025e0.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "fd50e5a5a03dd133ddb25b254afe55c273ad51b5b35df7e73fc2689c4b343058", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "35ca05ee88faf3ddb22a52647dc0f21348cd9633c251cf5a864d7752f9cfe125", "generation_seconds_batch": 16.132118637906387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7262609004974365, "target_z": 0.01855467321583921, "ce": 6.125868320465088, "genuineness": 4.482243061065674, "identity": 0.48814061284065247, "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_CART_reg/fd50e5a5a03dd133ddb25b254afe55c273ad51b5b35df7e73fc2689c4b343058.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-1", "key": "fd88b3a2b03000989a6eb6c63509c0f804c9750b8f8e02790107a54e480c981f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond naast het oude raam.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "e525dec8665cadfd3c3249d60880b9df93adcbc3d4f30c39fcd4b9f8d7611e56", "generation_seconds_batch": 17.01352252997458, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.567298412322998, "target_z": 0.6328124619089046, "ce": 5.816103935241699, "genuineness": 3.7165660858154297, "identity": -0.02769102342426777, "hypothesis": "Ik heb het een kleine doosje gevonden. Het stond... Het stond naast het oude raam. We kunnen er samen in kijken.", "wer": 0.16666666666666666, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/fd88b3a2b03000989a6eb6c63509c0f804c9750b8f8e02790107a54e480c981f.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": "fe2a47b334340d610c4d82828622e94fc2adf9f244de2b6bab14e5691c72c6d3", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Someone moved the little blue chair.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "959fb636424da863cfe5e3b756b461a701cea70afa450083ac83a220f25169d9", "generation_seconds_batch": 21.231656990014017, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.131250262260437, "target_z": -0.4160156209188112, "ce": 6.443574905395508, "genuineness": 4.507347106933594, "identity": 0.5910686254501343, "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_CART_reg/fe2a47b334340d610c4d82828622e94fc2adf9f244de2b6bab14e5691c72c6d3.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-0", "key": "fe4a240c44e23104b15fe101557dd344c6b3d2f9c994e0b2124e5a1f30168c92", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] La puerta sigue abierta.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "94c968854ed0d16ebc9595818796ac1ee382c3d0e654e77fd047515676357a52", "generation_seconds_batch": 19.24527993099764, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.583343744277954, "target_z": 0.6445312853703029, "ce": 5.629451274871826, "genuineness": 3.8922953605651855, "identity": 0.36345261335372925, "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_CART_reg/fe4a240c44e23104b15fe101557dd344c6b3d2f9c994e0b2124e5a1f30168c92.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": "fe4fdf4646fe2cb7bf22df04a3a38df5c75ec14cf2d240b4b5f677edd6e1c230", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I wrote the number on the envelope.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "94abcd6327662fa1369f7afe6d7d749a14369efa575b5f0b6d926dde068c6af5", "generation_seconds_batch": 15.302311728941277, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2408926486968994, "target_z": -0.3359375163247552, "ce": 6.213237762451172, "genuineness": 4.093128204345703, "identity": 0.30036604404449463, "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_CART_reg/fe4fdf4646fe2cb7bf22df04a3a38df5c75ec14cf2d240b4b5f677edd6e1c230.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "fe71c06edabf5d2a142f973e534eb0991644d221c89e10762011e2af5f67361f", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "be7df1f5d94da932990ded2e3b13e021877655f52488bf8046bc1186a469e39e", "generation_seconds_batch": 23.051225333008915, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.9523839950561523, "target_z": 0.9140624836752448, "ce": 5.696964263916016, "genuineness": 3.9048471450805664, "identity": 0.2732212543487549, "hypothesis": "¡Nick! ¡He encontrado la cajita! Estaba junto a la ventana vieja. Podemos mirar juntos lo que hay dentro.", "wer": 0.058823529411764705, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_CART_reg/fe71c06edabf5d2a142f973e534eb0991644d221c89e10762011e2af5f67361f.mp3" }, { "id": "vn_S_CART_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_CART_reg-en-1", "key": "febc928bb19b2fd0dcd0785df232fa30a79bea27b2ad95476590d7976a6ee7e0", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] I found the little box.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] It was beside the old window.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "48f1cc35a77ef5a5cd383f2f3965f5aef7d81daf1aef7331d78f7a2ae7aea448", "generation_seconds_batch": 18.970078411046416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6500461101531982, "target_z": -0.03710934643167842, "ce": 6.38797664642334, "genuineness": 4.394378662109375, "identity": 0.5691560506820679, "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_CART_reg/febc928bb19b2fd0dcd0785df232fa30a79bea27b2ad95476590d7976a6ee7e0.mp3" }, { "id": "vn_S_CART_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_CART_reg-fr-1", "key": "ff3264e059f8c10d8a9e7012d0c2c8a0a4423501d9bd9e01ec0b8db6dd42080e", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "a6799301b5f88d98a56318a7b1e47fd632f1eb37342b35f0f326f3f7d81d5181", "generation_seconds_batch": 15.080126675078645, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3640589714050293, "target_z": 0.4843750761821909, "ce": 5.903473377227783, "genuineness": 4.21237325668335, "identity": 0.560965359210968, "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_CART_reg/ff3264e059f8c10d8a9e7012d0c2c8a0a4423501d9bd9e01ec0b8db6dd42080e.mp3" }, { "id": "vn_S_CART_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_CART_reg-nl-0", "key": "ff652085cd64da91a0ab3d4ffe82a3a3cfec88b31edfb51767b733662894ef98", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] De deur staat nog open.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "e124d1a912c39a6b0e81f65e1ff6e6f11b54fec22f51151f4dc25641f734f812", "generation_seconds_batch": 15.758643896086141, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.884039044380188, "target_z": 0.13378907406336826, "ce": 6.197352409362793, "genuineness": 4.1998209953308105, "identity": 0.7238019704818726, "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_CART_reg/ff652085cd64da91a0ab3d4ffe82a3a3cfec88b31edfb51767b733662894ef98.mp3" }, { "id": "vn_S_CART_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_CART_reg-de-0", "key": "ff813531c2b84de188e49d3181dbb72813b4e363bf83a3c374e08bb540189e57", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Die Tür ist noch offen.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e46d8e8afe6f5fa8bde788516db90a4eaf3fb0269974ea3dce1615dc0448e443", "generation_seconds_batch": 20.58554262900725, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9520785212516785, "target_z": -0.5468749891168299, "ce": 6.23706579208374, "genuineness": 4.149612903594971, "identity": 0.5856006741523743, "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_CART_reg/ff813531c2b84de188e49d3181dbb72813b4e363bf83a3c374e08bb540189e57.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": "ffb6444dead26b19de0f3acce280cbe838e373328d89e815121b63c17fdec615", "bucket": "vn_S_CART_reg", "phase": "heldout", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e15cf37394eb997c51cf4143ced5a4e3e6074ce86ba336a9084397965d161812", "generation_seconds_batch": 19.841244725044817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1686891317367554, "target_z": -0.38867187363960376, "ce": 6.133810997009277, "genuineness": 4.231201171875, "identity": 0.3848688006401062, "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_CART_reg/ffb6444dead26b19de0f3acce280cbe838e373328d89e815121b63c17fdec615.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "ffb9d45a37363eafd1ee5ea96c2bd610089d2546a8eedd883fea7637066ab10c", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "4528bfd68e9e394d5d1f889b4becfcdeeaf78538d4291d2979dc56582b851ec9", "generation_seconds_batch": 20.265794175909832, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.663417100906372, "target_z": -0.027343747279207468, "ce": 6.09409761428833, "genuineness": 4.507347106933594, "identity": 0.7721909284591675, "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_CART_reg/ffb9d45a37363eafd1ee5ea96c2bd610089d2546a8eedd883fea7637066ab10c.mp3" }, { "id": "vn_S_CART_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_CART_reg-es-1", "key": "ffc8787a452e92b16e26c46a44044220baff8a527aad5bb567425f9a8395380a", "bucket": "vn_S_CART_reg", "phase": "grid", "candidate": "manual-cff870ad7de8", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Embrace a highly elastic vocal range, treating every word as a potential peak or valley in an exaggerated emotional landscape.\nSCRIPT:\n(Start with a sudden, bright upward lilt, making the initial sound feel slightly too bouncy for the context.) [3.2 seconds duration] He encontrado la cajita.\n(Maintain rapid, almost frantic pitch shifts between syllables, creating a constant sense of playful instability.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(End the delivery with a sudden, slightly clipped high note that snaps the energy off abruptly.) [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": "87a8e46c040e565c0f091678a185d4d3952d6e7bc785f1afed4c6ce8a1a89a31", "generation_seconds_batch": 24.308702151989564, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.856112480163574, "target_z": 0.8437498911682987, "ce": 5.80021858215332, "genuineness": 4.093128204345703, "identity": 0.4926985204219818, "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_CART_reg/ffc8787a452e92b16e26c46a44044220baff8a527aad5bb567425f9a8395380a.mp3" }, { "key": "original-vn_S_CART_reg-de-0-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "d748d647f678869b259699befed0906848a3ed80b59bdd7bce143af839e38a2d", "task_id": "vn_S_CART_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0323046445846558, "target_z": -0.4882812636039627, "ce": 6.292664527893066, "genuineness": 4.469690799713135, "identity": 0.4877336919307709, "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_CART_reg/de/a7e5bbcd0d186141a8cfc77e.mp3" }, { "key": "original-vn_S_CART_reg-de-0-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "e906fe2acea0e7ae2931b8c68c3b4c2135abb5d293a0d8d746ad6af47dc8fd8e", "task_id": "vn_S_CART_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.47606992721557617, "target_z": -0.8945312853703029, "ce": 6.372091293334961, "genuineness": 4.381826400756836, "identity": 0.49315929412841797, "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_CART_reg/de/b44288546155097dfb1bf0a8.mp3" }, { "key": "original-vn_S_CART_reg-de-0-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "23c5c6fb29a4fd69eab060c102586d6788af19b33a10ccb411cb62a7302927d7", "task_id": "vn_S_CART_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.9360332489013672, "target_z": -0.5585937690455477, "ce": 6.348263263702393, "genuineness": 4.111956596374512, "identity": 0.4297097325325012, "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_CART_reg/de/739d647b0f667b080d592742.mp3" }, { "key": "original-vn_S_CART_reg-de-0-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "e859d6230574219ce5e8d6b8aa6dfe12dfd740f42e51585cfc79e422154d4811", "task_id": "vn_S_CART_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.6204770803451538, "target_z": -0.7890624836752448, "ce": 6.3641486167907715, "genuineness": 4.381826400756836, "identity": 0.5824957489967346, "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_CART_reg/de/5a7284aa1e70ea9bcc72e058.mp3" }, { "key": "original-vn_S_CART_reg-de-0-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "8f000b3f4a9fa5f2fc4e9bce1b31887c3985f4ade2a0d5d17249c81419bc60ae", "task_id": "vn_S_CART_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.968123733997345, "target_z": -0.5351562527207925, "ce": 6.24500846862793, "genuineness": 4.6203155517578125, "identity": 0.4094308018684387, "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_CART_reg/de/31bcfac7e81036c6579110b6.mp3" }, { "key": "original-vn_S_CART_reg-de-0-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "2c1cca1d50312f165b071bccefb3bd8052e729448eb0d4959f63b6fc20a631c4", "task_id": "vn_S_CART_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.6365222930908203, "target_z": -0.7773437472792075, "ce": 6.23706579208374, "genuineness": 4.369274139404297, "identity": 0.7785303592681885, "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_CART_reg/de/cbd1f75387ea41bc19678f7e.mp3" }, { "key": "original-vn_S_CART_reg-de-0-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "e577985301685946fe22a10ff7a6ad10dd908f00eeb8631d18e522552254f7c7", "task_id": "vn_S_CART_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.4493279457092285, "target_z": -0.9140624836752448, "ce": 6.324435234069824, "genuineness": 4.469690799713135, "identity": 0.6414807438850403, "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_CART_reg/de/6baae280aad29855b0195a8a.mp3" }, { "key": "original-vn_S_CART_reg-de-0-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "79ab06cb2c4aca41ca678af9c4a3011aaee7934fd9300b63fb0efaf00dca8986", "task_id": "vn_S_CART_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.9360332489013672, "target_z": -0.5585937690455477, "ce": 6.205295085906982, "genuineness": 4.419482707977295, "identity": 0.4400370121002197, "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_CART_reg/de/04c1473aa6d30eb58638e905.mp3" }, { "key": "original-vn_S_CART_reg-de-1-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "b426abad4e52dbed67d469f70d9f859cdbf30f32142ddce04595b0c246e0a3f2", "task_id": "vn_S_CART_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.09633278846740723, "target_z": -1.1718750326495104, "ce": 6.427689552307129, "genuineness": 4.20609712600708, "identity": 0.5624316930770874, "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_CART_reg/de/e4baa54a6370bdcd7215efee.mp3" }, { "key": "original-vn_S_CART_reg-de-1-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "6e4c7dc635ac7f0f9adf17e64e3203ef3cdb256ac264b5646a9d90a26bbd4f4b", "task_id": "vn_S_CART_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.2674819231033325, "target_z": -1.0468750326495104, "ce": 6.356205940246582, "genuineness": 4.15588903427124, "identity": 0.4401434361934662, "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_CART_reg/de/c80d664df74d485dc66b753f.mp3" }, { "key": "original-vn_S_CART_reg-de-1-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "b1c0dd3420f99ea1902bfb2965834dcfe4d2420d7034cf774dd2300754680272", "task_id": "vn_S_CART_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.32096612453460693, "target_z": -1.0078124619089046, "ce": 6.507116317749023, "genuineness": 4.381826400756836, "identity": 0.5604754686355591, "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_CART_reg/de/977438c39ab1cedbccd6cb11.mp3" }, { "key": "original-vn_S_CART_reg-de-1-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "fb9eab83d590b27f6e55ba16326dd5656994d631f0702b1ae396d79595db06f6", "task_id": "vn_S_CART_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.845110297203064, "target_z": -0.625, "ce": 6.348263263702393, "genuineness": 4.218649387359619, "identity": 0.39662426710128784, "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_CART_reg/de/112dc47cd18538bbf456353a.mp3" }, { "key": "original-vn_S_CART_reg-de-1-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "5b21e814e571fbe1e82b2979b0dfa9c3e3f0def3291947b2cece70434269a4ef", "task_id": "vn_S_CART_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.2033010721206665, "target_z": -1.0937499782336597, "ce": 6.38003396987915, "genuineness": 4.281409740447998, "identity": 0.6472066640853882, "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_CART_reg/de/190a91205c95e707ec570d6a.mp3" }, { "key": "original-vn_S_CART_reg-de-1-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "d0b59b10e71a74aec5e05418bd9b53da4230e4edba670d6952fde49f69e50ff5", "task_id": "vn_S_CART_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.4172375202178955, "target_z": -0.9374999564673195, "ce": 6.38003396987915, "genuineness": 4.15588903427124, "identity": 0.6714238524436951, "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_CART_reg/de/37106c638ea90a653dc81fdf.mp3" }, { "key": "original-vn_S_CART_reg-de-1-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "cd21c7d44b587bfbae622930813b1a81bc8e77dce334b6fdeb7fe8a06007f7e5", "task_id": "vn_S_CART_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.5455993413925171, "target_z": -0.8437499782336597, "ce": 6.316492557525635, "genuineness": 4.042920112609863, "identity": 0.3676663637161255, "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_CART_reg/de/7b3ff045edf13ac6297174b9.mp3" }, { "key": "original-vn_S_CART_reg-de-1-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "636cd49f6b379993bf91812fbd62979e4b7471b405aafbefb1c7b8fbd2f9252e", "task_id": "vn_S_CART_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.4814183712005615, "target_z": -0.8906250108831701, "ce": 6.332377910614014, "genuineness": 4.20609712600708, "identity": 0.5917460918426514, "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_CART_reg/de/94aeed784e31fe19bcc3f830.mp3" }, { "key": "original-vn_S_CART_reg-en-0-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "4fbe18d524174948b084323b7259b0130e1be057f9d4d796d403083781e2eee3", "task_id": "vn_S_CART_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.9360332489013672, "target_z": -0.5585937690455477, "ce": 6.213237762451172, "genuineness": 3.948779582977295, "identity": 0.2501938045024872, "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_CART_reg/en/5048b07859c5fc20abc8e7a7.mp3" }, { "key": "original-vn_S_CART_reg-en-0-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "4314a00f165cac27339cae0a40cb3d9c1f83ad678ddca256fcc0789dbdc79df5", "task_id": "vn_S_CART_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.9734721183776855, "target_z": -0.5312500217663403, "ce": 6.197352409362793, "genuineness": 3.7918784618377686, "identity": 0.5101268887519836, "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_CART_reg/en/38789225f487cdd8aee589fd.mp3" }, { "key": "original-vn_S_CART_reg-en-0-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "1016dcaa6ed12cfdeabb1491f75441332c26b696045ce954c76bfbcb74e9a244", "task_id": "vn_S_CART_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0563725233078003, "target_z": -0.47070311547722615, "ce": 6.213237762451172, "genuineness": 3.8420867919921875, "identity": 0.3752185106277466, "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_CART_reg/en/f904698a0dca642d68b434bb.mp3" }, { "key": "original-vn_S_CART_reg-en-0-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "9479a37118c74514fdce891a20606c44643034cbf851eadf9e744af085197479", "task_id": "vn_S_CART_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.7274452447891235, "target_z": -0.7109375163247552, "ce": 6.2767791748046875, "genuineness": 4.068024158477783, "identity": 0.2978707551956177, "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_CART_reg/en/aad3007b89768a6521d51357.mp3" }, { "key": "original-vn_S_CART_reg-en-0-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "826c5ed92759fd147daba4ee3bb392a6e5620d5838946d2821bcfdf352db0b26", "task_id": "vn_S_CART_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0724177360534668, "target_z": -0.4589843790811888, "ce": 6.260893821716309, "genuineness": 3.8169827461242676, "identity": 0.393947571516037, "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_CART_reg/en/8a0efdd8172ec99f2bff7c8b.mp3" }, { "key": "original-vn_S_CART_reg-en-0-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "3ec7f37cae38d7f1191b90676b5c16d7b1383219387aa5bed9d80dfed16c362a", "task_id": "vn_S_CART_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.059046745300293, "target_z": -0.46874997823365977, "ce": 6.2767791748046875, "genuineness": 4.068024158477783, "identity": 0.30764180421829224, "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_CART_reg/en/84311bc42ed485bb664e7ed7.mp3" }, { "key": "original-vn_S_CART_reg-en-0-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "85131790c107bb3bdb807cab93ba3dac3bb2febad4c71998f1bfe965beb46a51", "task_id": "vn_S_CART_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.5562961101531982, "target_z": -0.8359375163247552, "ce": 6.308549880981445, "genuineness": 4.20609712600708, "identity": 0.35694268345832825, "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_CART_reg/en/e4ec555039e52b724da42e25.mp3" }, { "key": "original-vn_S_CART_reg-en-0-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "d4c21ee8f19fa563c4ddcaaa5909ff2e7767aa5ed784bad058a369980c94f861", "task_id": "vn_S_CART_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Tell me when you are ready.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2890284061431885, "target_z": -0.30078122007128216, "ce": 6.23706579208374, "genuineness": 4.055472373962402, "identity": 0.3354784846305847, "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_CART_reg/en/f06285bc9990d933ca69b6a2.mp3" }, { "key": "original-vn_S_CART_reg-en-1-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "5bd9672b794c7f6f10a4e5449552897b36e6b35d982545bddd6743ef29f55d6e", "task_id": "vn_S_CART_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.9199880361557007, "target_z": -0.5703125054415851, "ce": 6.268836498260498, "genuineness": 4.344170093536377, "identity": 0.3367213308811188, "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_CART_reg/en/386623c484187d7ca71af1dc.mp3" }, { "key": "original-vn_S_CART_reg-en-1-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "bd989eb2b979bca006934d8dfbf2c1ddef2ae3ae0a8f1cacd76704b055da609e", "task_id": "vn_S_CART_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3585577011108398, "target_z": -0.25, "ce": 6.340320587158203, "genuineness": 4.293961524963379, "identity": 0.3960813879966736, "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_CART_reg/en/ffe994ee8bc3f95f94bb7f49.mp3" }, { "key": "original-vn_S_CART_reg-en-1-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "64635d5e2e2183db084d496dfe83f52d530724571249e6a0dcfcc752c34976c1", "task_id": "vn_S_CART_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We can look inside together.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1633408069610596, "target_z": -0.3925780610613755, "ce": 6.300607204437256, "genuineness": 4.457139015197754, "identity": 0.3095269799232483, "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_CART_reg/en/18f0d40490d011e3994df065.mp3" }, { "key": "original-vn_S_CART_reg-en-1-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "832866a28d5440f3e3ce45007622d7c2525fd00c0be815b41e7032bf036f11a8", "task_id": "vn_S_CART_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.5883865356445312, "target_z": -0.8125000435326805, "ce": 6.348263263702393, "genuineness": 4.3127899169921875, "identity": 0.4591948986053467, "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_CART_reg/en/c43be328ebb89208adfb394b.mp3" }, { "key": "original-vn_S_CART_reg-en-1-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "75fa5672b4333c5f0a309c1e0fa2c604a58c3d3199ceacfbf6d2084762a2826b", "task_id": "vn_S_CART_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.9092912077903748, "target_z": -0.5781250108831701, "ce": 6.268836498260498, "genuineness": 4.3504462242126465, "identity": 0.37269127368927, "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_CART_reg/en/ea8d2d0fba28c1240ed96340.mp3" }, { "key": "original-vn_S_CART_reg-en-1-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "8b0a9872fd17ecd319a8f8e179c229f149b4f7c89ed260e79a9b19ec1cd7b5b4", "task_id": "vn_S_CART_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.8718523383140564, "target_z": -0.6054687581623776, "ce": 6.324435234069824, "genuineness": 4.482243061065674, "identity": 0.3985831141471863, "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_CART_reg/en/095e25274442e5f2f65d80ba.mp3" }, { "key": "original-vn_S_CART_reg-en-1-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "9f23bba1822e546793287908f8a01e4a219bcd6e67125fb8ec69748f3967f0b7", "task_id": "vn_S_CART_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.6044318675994873, "target_z": -0.8007812200712822, "ce": 6.427689552307129, "genuineness": 4.3504462242126465, "identity": 0.36727577447891235, "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_CART_reg/en/cd6a17591fda15c877e03950.mp3" }, { "key": "original-vn_S_CART_reg-en-1-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "0cbb40bb373cb02930fcee403a1b5dff55c3c327537b83218b3cf1b0a88e6728", "task_id": "vn_S_CART_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.6846579313278198, "target_z": -0.7421875380910954, "ce": 6.4515180587768555, "genuineness": 4.494794845581055, "identity": 0.19759228825569153, "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_CART_reg/en/5cc09daeddcd6d0e5caf6d0b.mp3" }, { "key": "original-vn_S_CART_reg-es-0-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "61227e482c052a06e13c11ea4ba5266d6e9a40b43fa82adfddc40dd93f3dc9b8", "task_id": "vn_S_CART_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.7327936887741089, "target_z": -0.7070312418376224, "ce": 5.895530700683594, "genuineness": 4.293961524963379, "identity": 0.5835145711898804, "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_CART_reg/es/00686781c7eb201d79a6cab1.mp3" }, { "key": "original-vn_S_CART_reg-es-0-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "67005e884ab1105680eec52762e6f01879ad88e683b73b305b020e8514201ecf", "task_id": "vn_S_CART_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.7274452447891235, "target_z": -0.7109375163247552, "ce": 5.990842819213867, "genuineness": 4.030367851257324, "identity": 0.5160273313522339, "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_CART_reg/es/78647dd25422ce941953381a.mp3" }, { "key": "original-vn_S_CART_reg-es-0-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "0dd0599664d7ab82ce03de49320175587333fe1d46f10233090c34e05f66557e", "task_id": "vn_S_CART_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3318157196044922, "target_z": -0.2695311983049419, "ce": 6.062326908111572, "genuineness": 4.055472373962402, "identity": 0.5040571093559265, "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_CART_reg/es/e244af48c9cb6afebaeafbbd.mp3" }, { "key": "original-vn_S_CART_reg-es-0-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "3c4c1d18c68c1fbd0c17cd2964492227016e0db6350751a05334b7eb6d7d6a56", "task_id": "vn_S_CART_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1606664657592773, "target_z": -0.3945312853703029, "ce": 5.621508598327637, "genuineness": 4.13078498840332, "identity": 0.5219277143478394, "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/audio/vn_S_CART_reg/es/10b5e6657e686d9fcd1a2b7c.mp3" }, { "key": "original-vn_S_CART_reg-es-0-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "f798861603f1fd0627d4cfc7b23ed7f77df337929da50462c5d239d1e42a0f07", "task_id": "vn_S_CART_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6888220310211182, "target_z": -0.00878907406336826, "ce": 5.990842819213867, "genuineness": 4.13078498840332, "identity": 0.5160629153251648, "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_CART_reg/es/f050d7bb70e4a6d6add07537.mp3" }, { "key": "original-vn_S_CART_reg-es-0-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "74e021ed39d0d15a2cf019d54cd83388df4cb12dceb9dace0766d223a5581d5b", "task_id": "vn_S_CART_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5644714832305908, "target_z": -0.09960938996435892, "ce": 5.728734970092773, "genuineness": 4.18099308013916, "identity": 0.6415129899978638, "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_CART_reg/es/100115367d9583b8f4a06a94.mp3" }, { "key": "original-vn_S_CART_reg-es-0-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "154742846a707ad1fcdf3eab2576344c95ebc8281d679adae392d11f36105b70", "task_id": "vn_S_CART_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.139272928237915, "target_z": -0.410156209188112, "ce": 6.070269584655762, "genuineness": 3.961331605911255, "identity": 0.49801474809646606, "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_CART_reg/es/9cd9afef38d20031c283abc9.mp3" }, { "key": "original-vn_S_CART_reg-es-0-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "8de5192787a80b6253fa3a2b1b0f342cb70a12aa472ce25dcfc940cf6fbf2bf1", "task_id": "vn_S_CART_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dime cuándo estás listo.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2622863054275513, "target_z": -0.32031250544158507, "ce": 6.0146708488464355, "genuineness": 4.18099308013916, "identity": 0.658159077167511, "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_CART_reg/es/a1755753e4e42584975b75af.mp3" }, { "key": "original-vn_S_CART_reg-es-1-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "1fd072c6dba5c41ecaf68e303bf8e1cd57631031ddd0b477d1e28747216c63ca", "task_id": "vn_S_CART_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0376530885696411, "target_z": -0.48437498911682986, "ce": 6.173524379730225, "genuineness": 4.607763767242432, "identity": 0.5893056392669678, "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_CART_reg/es/dacc3ef8d948cb08bb3e4432.mp3" }, { "key": "original-vn_S_CART_reg-es-1-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "fc4021e596f9d1e967ea5046bdc2658dea6c32a536fb14b776982ee76c05e47e", "task_id": "vn_S_CART_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.7488389015197754, "target_z": -0.6953125054415851, "ce": 6.165581703186035, "genuineness": 4.457139015197754, "identity": 0.6972732543945312, "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_CART_reg/es/4bb30f1ba8fc10e37faf992c.mp3" }, { "key": "original-vn_S_CART_reg-es-1-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "293f6d9ae0749402fb24f64c7b89ea396b55eb82efebf1de8fecbc4bed2e4e7c", "task_id": "vn_S_CART_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3478608131408691, "target_z": -0.2578125489742656, "ce": 6.221180438995361, "genuineness": 4.457139015197754, "identity": 0.5856454372406006, "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_CART_reg/es/1092ccacee5953a8865e3af6.mp3" }, { "key": "original-vn_S_CART_reg-es-1-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "ae15ce5403f848167404c9bed8e0277c341fe79b04ed1a815522533c4cfc1163", "task_id": "vn_S_CART_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.6151286363601685, "target_z": -0.7929687581623776, "ce": 5.974957466125488, "genuineness": 4.381826400756836, "identity": 0.6163706183433533, "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_CART_reg/es/9416677c8aa08c0174d9d638.mp3" }, { "key": "original-vn_S_CART_reg-es-1-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "5954a70338b76488b3b2a769a24f00fac78597b3d90f2dabe5bd89b4ed6ecda2", "task_id": "vn_S_CART_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.8397619128227234, "target_z": -0.6289062309544523, "ce": 6.157639026641846, "genuineness": 4.369274139404297, "identity": 0.6600710153579712, "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_CART_reg/es/4cbfaa4a43dbed6e108627a6.mp3" }, { "key": "original-vn_S_CART_reg-es-1-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "0a23881266f265edd6851ee0104e21a1dab184964bfdd418d5197d465588cb70", "task_id": "vn_S_CART_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.722096860408783, "target_z": -0.7148437472792075, "ce": 6.125868320465088, "genuineness": 4.13078498840332, "identity": 0.7037293910980225, "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_CART_reg/es/0b70d53c110d09be8ada63b7.mp3" }, { "key": "original-vn_S_CART_reg-es-1-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "a195f057ba4fa9fddac4137df689a39542e10eef73d76513f33e8a3ac0845018", "task_id": "vn_S_CART_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3104219436645508, "target_z": -0.28515629625347305, "ce": 6.022613525390625, "genuineness": 4.419482707977295, "identity": 0.6581845283508301, "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_CART_reg/es/62ace18f6aed136548da3f05.mp3" }, { "key": "original-vn_S_CART_reg-es-1-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "682a63111288c9ac21e0b5d24ce8dad18ef642cf9734638658b36f9e7168bf46", "task_id": "vn_S_CART_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1365985870361328, "target_z": -0.41210943349703943, "ce": 6.125868320465088, "genuineness": 4.545003414154053, "identity": 0.7271486520767212, "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_CART_reg/es/69c356f9f9b9b0c64a8ff67e.mp3" }, { "key": "original-vn_S_CART_reg-fr-0-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "dd637bb7a7c396506183c322ab38570de394c9d055532eb3a96d281c7e08aa3a", "task_id": "vn_S_CART_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.834413468837738, "target_z": -0.6328125054415851, "ce": 6.268836498260498, "genuineness": 4.507347106933594, "identity": 0.6856623291969299, "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_CART_reg/fr/503d00c45c4b07c331d2cecb.mp3" }, { "key": "original-vn_S_CART_reg-fr-0-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "16d43f51f2d0328487aecef635a229add513a292cee6ff48af5bfc3987b5d4d7", "task_id": "vn_S_CART_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.7488389015197754, "target_z": -0.6953125054415851, "ce": 6.109982967376709, "genuineness": 4.105680465698242, "identity": 0.6303784847259521, "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_CART_reg/fr/9e6716b257ba0d6ec5d55c65.mp3" }, { "key": "original-vn_S_CART_reg-fr-0-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "898609c96498ad20f3b2d3a5692871b1561318564ecb3d66d514ef802367fd1c", "task_id": "vn_S_CART_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.6739611625671387, "target_z": -0.75, "ce": 6.149696350097656, "genuineness": 4.268857479095459, "identity": 0.7282323241233826, "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_CART_reg/fr/348b66aaaecbc12ce4450a64.mp3" }, { "key": "original-vn_S_CART_reg-fr-0-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "b9a9d8e0dc0dd666d0cac3bd5ea08ec20dc545479feea974e19dbd0fe865c0a7", "task_id": "vn_S_CART_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.7702325582504272, "target_z": -0.6796874945584149, "ce": 6.181467056274414, "genuineness": 4.381826400756836, "identity": 0.7398580312728882, "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_CART_reg/fr/2c8106eff9535bcbf12f592b.mp3" }, { "key": "original-vn_S_CART_reg-fr-0-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "942abd86bcffcbf76acc99717fd63808a567e9fa19768055fa90f1c7620160b0", "task_id": "vn_S_CART_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.9360332489013672, "target_z": -0.5585937690455477, "ce": 6.133810997009277, "genuineness": 4.042920112609863, "identity": 0.8253496885299683, "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_CART_reg/fr/f756122db193e165d94d572e.mp3" }, { "key": "original-vn_S_CART_reg-fr-0-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "41f6544e144d1720b03dd31d3874f08692611685655b5c3cc5d9054400b05ba1", "task_id": "vn_S_CART_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.203453779220581, "target_z": -0.3632812636039627, "ce": 6.308549880981445, "genuineness": 4.2751336097717285, "identity": 0.7338279485702515, "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_CART_reg/fr/daab1fdb95d0ca4df8fe7a0f.mp3" }, { "key": "original-vn_S_CART_reg-fr-0-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "97ee7aafe1547cda8e86953d8ebf8fdf55173d29f0c4960f63d79a0e4de191ef", "task_id": "vn_S_CART_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.6525675058364868, "target_z": -0.7656250108831701, "ce": 6.181467056274414, "genuineness": 4.337893962860107, "identity": 0.8114235401153564, "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_CART_reg/fr/8a94ff29298dae3e903a543e.mp3" }, { "key": "original-vn_S_CART_reg-fr-0-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "ed3468d80e592112476076111b63449d49ff289c565e87b556da5a7197ce16cb", "task_id": "vn_S_CART_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1125308275222778, "target_z": -0.42968749455841493, "ce": 6.09409761428833, "genuineness": 4.381826400756836, "identity": 0.45288902521133423, "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_CART_reg/fr/730fd67b33393c26bab1ebe0.mp3" }, { "key": "original-vn_S_CART_reg-fr-1-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "870a7cbbb9ceba7a24235e2cf97cfeab403348bcd81a7bc4199f8a9a959470e1", "task_id": "vn_S_CART_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.4172375202178955, "target_z": -0.9374999564673195, "ce": 6.165581703186035, "genuineness": 4.68307638168335, "identity": 0.719978928565979, "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_CART_reg/fr/690c568999a950ba9d24edca.mp3" }, { "key": "original-vn_S_CART_reg-fr-1-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "21b2f245a7e8a94f8ee8382a3a428af500f5061d65c0d7b3f30c8db40836f5bc", "task_id": "vn_S_CART_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.599083423614502, "target_z": -0.8046874945584149, "ce": 6.09409761428833, "genuineness": 4.582659721374512, "identity": 0.7811203002929688, "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_CART_reg/fr/af4864f24b40a41c378b9c96.mp3" }, { "key": "original-vn_S_CART_reg-fr-1-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "daedbf39e08dfe4433b39ae77fd8fb75c6fd9462b435c29a0b5c49999b24ca22", "task_id": "vn_S_CART_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.5081604719161987, "target_z": -0.8710937255128672, "ce": 6.23706579208374, "genuineness": 4.406930446624756, "identity": 0.8277310729026794, "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_CART_reg/fr/d3e2536d7d2ac54707c913fd.mp3" }, { "key": "original-vn_S_CART_reg-fr-1-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "c721f873cdf7613c8ce614b9cc5bf009e9f82374a7ee85136ef235058a2f1a8c", "task_id": "vn_S_CART_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.24608826637268066, "target_z": -1.0625000435326806, "ce": 6.109982967376709, "genuineness": 4.65797233581543, "identity": 0.7557435035705566, "hypothesis": "J'ai trouvé la la petite boîte. Elle était près de 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/audio/vn_S_CART_reg/fr/c0f728c97cc966b96a24dc95.mp3" }, { "key": "original-vn_S_CART_reg-fr-1-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "d316b8dc4b218560b697c6465a51ae8c066245f7071f4fb7ec4f8e2bdef2349b", "task_id": "vn_S_CART_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.6472190618515015, "target_z": -0.7695312853703029, "ce": 6.149696350097656, "genuineness": 4.7332844734191895, "identity": 0.714631199836731, "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_CART_reg/fr/334e21efa3cd28013419be6c.mp3" }, { "key": "original-vn_S_CART_reg-fr-1-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "efbc71122c1ee61b783ced14b6bc7fd2cde524459c78094cce7a186241606cac", "task_id": "vn_S_CART_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.18190741539001465, "target_z": -1.1093749891168299, "ce": 6.062326908111572, "genuineness": 4.7458367347717285, "identity": 0.7234485745429993, "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_CART_reg/fr/839d398261881b9d52d67452.mp3" }, { "key": "original-vn_S_CART_reg-fr-1-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.75, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "877322a8f8c949b3a64d1cd59ce829e19b3ccb6b6d84210fe3945673a5efa09b", "task_id": "vn_S_CART_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.24608826637268066, "target_z": -1.0625000435326806, "ce": 6.181467056274414, "genuineness": 4.632867813110352, "identity": 0.7707144021987915, "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/audio/vn_S_CART_reg/fr/4daf9e96b5bd6357a6d68ef5.mp3" }, { "key": "original-vn_S_CART_reg-fr-1-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "88b578f473c475b6ca902af76a69790065b8050fa83a6bb334ec875c13e94157", "task_id": "vn_S_CART_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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": 0.9360332489013672, "target_z": -0.5585937690455477, "ce": 6.149696350097656, "genuineness": 4.645420074462891, "identity": 0.8310994505882263, "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_CART_reg/fr/43491c4ba1edad3efa485156.mp3" }, { "key": "original-vn_S_CART_reg-nl-0-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "444a2d214cb2d54d07b9fe8330c08218ed46ab58090e2a16934f3a5aa38fcc06", "task_id": "vn_S_CART_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.2674819231033325, "target_z": -1.0468750326495104, "ce": 6.284721851348877, "genuineness": 4.425758361816406, "identity": 0.6756479740142822, "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_CART_reg/nl/dd32cbed1e7e929c36106c0f.mp3" }, { "key": "original-vn_S_CART_reg-nl-0-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "e7c85e36a63b92bdcefddb72ccddbdc0f8eefa96f02d39b819fa97d61adc3e70", "task_id": "vn_S_CART_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0349788665771484, "target_z": -0.48632812636039624, "ce": 6.221180438995361, "genuineness": 4.331617832183838, "identity": 0.5987673997879028, "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_CART_reg/nl/628b02f8210cd505c74cb288.mp3" }, { "key": "original-vn_S_CART_reg-nl-0-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "19e60b2149e4dd0841d41fbb1fbcf4c2ac20fda7a96e249de8c0ed1f017a5809", "task_id": "vn_S_CART_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5604602098464966, "target_z": -0.102539052297028, "ce": 6.125868320465088, "genuineness": 4.231201171875, "identity": 0.3158535957336426, "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_CART_reg/nl/3105d240aefed591118564b0.mp3" }, { "key": "original-vn_S_CART_reg-nl-0-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "965e7e3af53f50a1290b83088780ad91cdf4651b53e82fcea64065ea4c93d961", "task_id": "vn_S_CART_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.711400032043457, "target_z": -0.7226562527207925, "ce": 6.133810997009277, "genuineness": 4.243753433227539, "identity": 0.6031271815299988, "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_CART_reg/nl/8be7e86af464d12b37ab202f.mp3" }, { "key": "original-vn_S_CART_reg-nl-0-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "80f0c5f1cf69a216835b68a99a6c54b7fb959632e1c1934cbb79381bf332d442", "task_id": "vn_S_CART_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.40119218826293945, "target_z": -0.9492187799287178, "ce": 6.1020402908325195, "genuineness": 4.507347106933594, "identity": 0.6636478900909424, "hypothesis": "De deur staan nog open. Ik heb je jas op de stoel gelegd. Zeg maar wanneer je klaar bent.", "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_CART_reg/nl/e62da655bd70e1a3413450b7.mp3" }, { "key": "original-vn_S_CART_reg-nl-0-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "8ac5383ec6231140f6437949b44b798d6cd062090a996bcdcf6a90dfb7100302", "task_id": "vn_S_CART_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.10702967643737793, "target_z": -1.164062483675245, "ce": 5.935244083404541, "genuineness": 4.105680465698242, "identity": 0.6461365222930908, "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_CART_reg/nl/e93fffa5e506d302a7c2f28c.mp3" }, { "key": "original-vn_S_CART_reg-nl-0-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "64546eae051180418d0aa82e6371607fb38d13d8732904be33c58933d8d56e3d", "task_id": "vn_S_CART_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.711400032043457, "target_z": -0.7226562527207925, "ce": 6.205295085906982, "genuineness": 4.293961524963379, "identity": 0.7319997549057007, "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_CART_reg/nl/2a59c8f44f5c8b256bb9dc8f.mp3" }, { "key": "original-vn_S_CART_reg-nl-0-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "8d083c8758e598cefc6da178d0dcfd359b02e33d3678b255e1b729edab5465e8", "task_id": "vn_S_CART_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.566992998123169, "target_z": -0.8281249673504896, "ce": 6.086154937744141, "genuineness": 4.18726921081543, "identity": 0.7933383584022522, "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_CART_reg/nl/522126515295f1fb5f231603.mp3" }, { "key": "original-vn_S_CART_reg-nl-1-seed1701-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "2704df2896baa7e8295f870ae5ec69c47811096e0187cbcbf0b5eebbccb31fad", "task_id": "vn_S_CART_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3184446096420288, "target_z": -0.27929688452277385, "ce": 6.149696350097656, "genuineness": 4.362998008728027, "identity": 0.6658695340156555, "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_CART_reg/nl/be7ae6dda27fc6120f9df094.mp3" }, { "key": "original-vn_S_CART_reg-nl-1-seed1701-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "d6c00890044a437bdb50c5d69ea3df19105c65eb85a03aef516e9203be161e61", "task_id": "vn_S_CART_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5778424739837646, "target_z": -0.08984379081188798, "ce": 6.1894097328186035, "genuineness": 4.13078498840332, "identity": 0.6459472179412842, "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_CART_reg/nl/bc8f387b98310d713aa7bd2a.mp3" }, { "key": "original-vn_S_CART_reg-nl-1-seed1701-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "84ec367631830e7c029397c9dd15ab1d178c7d709670869001aca00deb6895ea", "task_id": "vn_S_CART_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5470892190933228, "target_z": -0.11230465144949896, "ce": 6.09409761428833, "genuineness": 4.306513786315918, "identity": 0.632536768913269, "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_CART_reg/nl/591e5c37295a429e601e330f.mp3" }, { "key": "original-vn_S_CART_reg-nl-1-seed1701-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "8ab2a1c07436a318cf2c253b540089c4b3d45eee5f7a0da9d3bfdaf14dcd77da", "task_id": "vn_S_CART_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9039428234100342, "target_z": -0.5820312418376224, "ce": 6.260893821716309, "genuineness": 4.582659721374512, "identity": 0.7917576432228088, "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_CART_reg/nl/441b8e824ebf27a6618cda7c.mp3" }, { "key": "original-vn_S_CART_reg-nl-1-seed1702-scale0.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "59af847683f41dba1a9330f3d020dfcf5333a3faa44cd3fe3af8270443823083", "task_id": "vn_S_CART_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.599083423614502, "target_z": -0.8046874945584149, "ce": 6.3641486167907715, "genuineness": 4.532451152801514, "identity": 0.6045889854431152, "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_CART_reg/nl/6964f9e6714705b5d3fe60ad.mp3" }, { "key": "original-vn_S_CART_reg-nl-1-seed1702-scale0.5", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "652a9f56ff6ab0d7d020f270858cbae1876248c43e456c83d9e6e5d43faff4d2", "task_id": "vn_S_CART_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7060516476631165, "target_z": -0.7265624836752448, "ce": 6.221180438995361, "genuineness": 4.168440818786621, "identity": 0.4794352650642395, "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_CART_reg/nl/ac90a2dc58a6a046973e81db.mp3" }, { "key": "original-vn_S_CART_reg-nl-1-seed1702-scale0.75", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "18151ea04086b1bd1b58e3e19e523a65c6f7798cdbaa1a65dc4a42151be09d6b", "task_id": "vn_S_CART_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8558071255683899, "target_z": -0.6171874945584149, "ce": 6.316492557525635, "genuineness": 4.356721878051758, "identity": 0.5731239318847656, "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_CART_reg/nl/ad2973658c54dbce8387bdf0.mp3" }, { "key": "original-vn_S_CART_reg-nl-1-seed1702-scale1.0", "bucket": "vn_S_CART_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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Cartoonish 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": "6a9d5336cca555ea7a302c43983fe73fdf57a84a60e7b8677a0db25280c637da", "mp3_sha256": "dca28a58c1b8d08771aa3d93a42ef5bf723a667fb24bb6279b7e2188f48ec0c6", "task_id": "vn_S_CART_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Cartoonish 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 Cartoonish 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 Cartoonish Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] We kunnen er samen in kijken.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.47072160243988037, "target_z": -0.8984374727920746, "ce": 6.1894097328186035, "genuineness": 4.381826400756836, "identity": 0.6880494356155396, "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_CART_reg/nl/b3a0ba4a62e3e7620d5919e7.mp3" } ], "source_standard_audio_unchanged": true }