{ "complete": true, "report": { "bucket": "vn_S_NEWS_reg", "label": "Strong Newsreader Style", "definition": { "id": "vn_S_NEWS_reg", "category": "talking_style", "head": "vn_S_NEWS_reg", "pole": "high", "label": "Strong Newsreader Style", "floor": 4.0, "canonical_dimension_name": "Newsreader Style" }, "recipe_status": "direct_manual_recipe", "manual_provenance": [ { "id": "rank4/style_s_news/full-v2-sealed/sealed-selected", "mission": "style_s_news", "format_status": "CANONICAL", "heldout": true, "qualified": true, "source_cfg": 3, "source_model": { "label": "Humaneness 1.5 + real-speech rank 4", "repo": "laion/Humaneness-Voice-Large-1.5-RealSpeech-LoRAs", "revision": "fed33e554f094fcdc89fc0124aed2f9d42018714", "rank": 4 }, "metrics": { "ce": 6.39989049911499, "genuineness": 4.075994892120361, "wer": 0.007596588802373007, "identity": 0.5917478692531586, "voiceclap": -0.12544025764334946, "fitness": 0.7460658465062594, "steering_v2": 0.8428301851233846, "raw_target": 1.2230304652452468, "target_cdf": 0.8428301851233846, "family_hit_all": null, "nominal_hit_all": null, "n": 100, "valid_fraction": 0.99, "by_arm": { "reference": { "n": 50, "ce": 6.393218669891358, "genuineness": 4.066894617080688, "wer": 0.005827956989247311, "identity": 0.5917478692531586, "fitness": 0.7294505668692182, "steering_v2": 0.8073957919725169 }, "none": { "n": 50, "ce": 6.406562328338623, "genuineness": 4.085095167160034, "wer": 0.009365220615498703, "identity": null, "fitness": 0.7626811261433006, "steering_v2": 0.8782645782742523 } } } } ], "selected_wording": { "id": "manual-59e65fa10d8f", "general": "Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.", "inline": [ "Start with a crisp, clear articulation that establishes immediate factual distance.", "Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.", "Preserve this highly controlled, precise delivery right up to the very last sound." ], "cfg": 3, "recipe": "P06", "strategy": "local-Gemma-mutation", "timing_syntax": "canonical", "canonical_labels": true, "neutral_event": "keep", "parent": "cfg-3", "origin": "exact_frozen_manual_recipe" }, "selection": { "manual_no_adapter": { "config": { "scale": 0.0, "cfg": 4.0 }, "n": 20, "target_z": 1.784179692042482, "target_raw": 1.4705521762371063, "ce": 6.218400478363037, "genuineness": 4.178482639789581, "identity": 0.569031224399805, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.05402777642011643, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, "manual_adapter": { "config": { "scale": 1.0, "cfg": 5.0 }, "n": 20, "target_z": 1.8779297184100043, "target_raw": 1.5222088932991027, "ce": 6.181864190101623, "genuineness": 4.1138393640518185, "identity": 0.5037363305687904, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.011267117410898208, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, "grid": [ { "config": { "scale": 0.0, "cfg": 4.0 }, "n": 20, "target_z": 1.784179692042482, "target_raw": 1.4705521762371063, "ce": 6.218400478363037, "genuineness": 4.178482639789581, "identity": 0.569031224399805, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.05402777642011643, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 3.0 }, "n": 20, "target_z": 1.7696288761745447, "target_raw": 1.4625346064567566, "ce": 6.198543763160705, "genuineness": 4.190720880031586, "identity": 0.5406344830989838, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": 0.02563103511929512, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 5.0 }, "n": 20, "target_z": 1.7340820267392099, "target_raw": 1.4429481208324433, "ce": 6.221180438995361, "genuineness": 4.177227449417114, "identity": 0.5584190264344215, "wer": 0.021111111111111112, "valid_fraction": 0.9, "identity_delta": 0.043415578454732894, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 2.5 }, "n": 20, "target_z": 1.5785156128726299, "target_raw": 1.3572302758693695, "ce": 6.200529479980469, "genuineness": 4.213000845909119, "identity": 0.52622299939394, "wer": 0.005263157894736842, "valid_fraction": 1, "identity_delta": 0.011219551414251327, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 2.0 }, "n": 20, "target_z": 1.3748046991414302, "target_raw": 1.2449845790863037, "ce": 6.204897952079773, "genuineness": 4.281409692764282, "identity": 0.5641372963786125, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.049133848398923874, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.0 }, "n": 20, "target_z": 1.351562509634287, "target_raw": 1.232178020477295, "ce": 6.211252069473266, "genuineness": 4.2572469234466555, "identity": 0.591065575927496, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.07606212794780731, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.5 }, "n": 20, "target_z": 1.3166015384354102, "target_raw": 1.2129143565893172, "ce": 6.183452701568603, "genuineness": 4.302434372901916, "identity": 0.5255330711603164, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": 0.010529623180627824, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.0, "cfg": 1.25 }, "n": 20, "target_z": 1.300488310290192, "target_raw": 1.2040358901023864, "ce": 6.211649179458618, "genuineness": 4.278271675109863, "identity": 0.5167611928656697, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.0017577448859810829, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 5.0 }, "n": 20, "target_z": 1.8779297184100043, "target_raw": 1.5222088932991027, "ce": 6.181864190101623, "genuineness": 4.1138393640518185, "identity": 0.5037363305687904, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.011267117410898208, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 5.0 }, "n": 20, "target_z": 1.8583984750863962, "target_raw": 1.5114470839500427, "ce": 6.190204000473022, "genuineness": 4.105994379520416, "identity": 0.47398021519184114, "wer": 0.0, "valid_fraction": 1, "identity_delta": -0.041023232787847516, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 4.0 }, "n": 20, "target_z": 1.8292968974377475, "target_raw": 1.4954119741916656, "ce": 6.175510025024414, "genuineness": 4.104425382614136, "identity": 0.5444392487406731, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.02943580076098442, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 5.0 }, "n": 20, "target_z": 1.8153320030760487, "target_raw": 1.4877172499895095, "ce": 6.189409685134888, "genuineness": 4.10505291223526, "identity": 0.5345263257622719, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.019522877782583235, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 3.0 }, "n": 20, "target_z": 1.7986327885199214, "target_raw": 1.4785159021615981, "ce": 6.1806727886199955, "genuineness": 4.128274297714233, "identity": 0.5929230831563472, "wer": 0.003125, "valid_fraction": 1, "identity_delta": 0.07791963517665863, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 3.0 }, "n": 20, "target_z": 1.7898437495774435, "target_raw": 1.4736730992794036, "ce": 6.209663534164429, "genuineness": 4.1844448208808895, "identity": 0.5103644773364067, "wer": 0.003125, "valid_fraction": 1, "identity_delta": -0.0046389706432819365, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 4.0 }, "n": 20, "target_z": 1.7629883170510952, "target_raw": 1.4588756263256073, "ce": 6.200926566123963, "genuineness": 4.162164890766144, "identity": 0.5664601892232894, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.05145674124360085, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 2.5 }, "n": 20, "target_z": 1.725195311676015, "target_raw": 1.4380514979362489, "ce": 6.219194769859314, "genuineness": 4.1850724697113035, "identity": 0.5634886041283608, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.048485156148672104, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 2.5 }, "n": 20, "target_z": 1.6634765732540617, "target_raw": 1.4040441751480102, "ce": 6.199735164642334, "genuineness": 4.222728681564331, "identity": 0.5463913455605507, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": 0.031387897580862044, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 2.0 }, "n": 20, "target_z": 1.6503417879960325, "target_raw": 1.3968068450689315, "ce": 6.181268501281738, "genuineness": 4.25756071805954, "identity": 0.5936692595481873, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.07866581156849861, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.5 }, "n": 20, "target_z": 1.5907226399498848, "target_raw": 1.363956406712532, "ce": 6.227137446403503, "genuineness": 4.236849844455719, "identity": 0.5295040905475616, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.014500642567873001, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 3.0 }, "n": 20, "target_z": 1.5770507631328923, "target_raw": 1.3564231365919113, "ce": 6.181864142417908, "genuineness": 4.189465713500977, "identity": 0.5621755182743072, "wer": 0.03368055555555556, "valid_fraction": 0.95, "identity_delta": 0.047172070294618604, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 4.0 }, "n": 20, "target_z": 1.5404297062403784, "target_raw": 1.3362447574734688, "ce": 6.198940944671631, "genuineness": 4.1618511319160465, "identity": 0.5317395824939013, "wer": 0.005756578947368421, "valid_fraction": 1, "identity_delta": 0.01673613451421261, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.25 }, "n": 20, "target_z": 1.5267578564669986, "target_raw": 1.3287115022540092, "ce": 6.195366740226746, "genuineness": 4.2349670171737674, "identity": 0.566770114004612, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.05176666602492332, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 2.0 }, "n": 20, "target_z": 1.5106445228516905, "target_raw": 1.3198329776525497, "ce": 6.211649179458618, "genuineness": 4.2519122838974, "identity": 0.5457712553441525, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.030767807364463808, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 2.5 }, "n": 20, "target_z": 1.4847656460010557, "target_raw": 1.3055735915899276, "ce": 6.223563218116761, "genuineness": 4.268543803691864, "identity": 0.5592631414532662, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.0442596934735775, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.25 }, "n": 20, "target_z": 1.3896484406057898, "target_raw": 1.2531635522842408, "ce": 6.250568318367004, "genuineness": 4.30117918252945, "identity": 0.513204725459218, "wer": 0.0027777777777777775, "valid_fraction": 1, "identity_delta": -0.0017987225204706193, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.5 }, "n": 20, "target_z": 1.3563964797004693, "target_raw": 1.234841561317444, "ce": 6.222768926620484, "genuineness": 4.300551581382751, "identity": 0.569177208840847, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.05417376086115837, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.25 }, "n": 20, "target_z": 1.315136702217479, "target_raw": 1.2121072247624398, "ce": 6.245405578613282, "genuineness": 4.317496955394745, "identity": 0.5438811019062996, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.028877653926610947, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.5, "cfg": 1.0 }, "n": 20, "target_z": 1.252050790176505, "target_raw": 1.177346582710743, "ce": 6.213237714767456, "genuineness": 4.341345846652985, "identity": 0.5348234787583351, "wer": 0.021052631578947368, "valid_fraction": 0.95, "identity_delta": 0.01982003077864647, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.0 }, "n": 20, "target_z": 1.2333007692177052, "target_raw": 1.1670152306556703, "ce": 6.2243575096130375, "genuineness": 4.279840648174286, "identity": 0.5408439941704273, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.025840546190738677, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 2.0 }, "n": 20, "target_z": 1.222412123673254, "target_raw": 1.1610155344009399, "ce": 6.221577572822571, "genuineness": 4.326597118377686, "identity": 0.5437135189771652, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.028710070997476578, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 0.75, "cfg": 1.5 }, "n": 20, "target_z": 1.2170898352037958, "target_raw": 1.1580829277634621, "ce": 6.224357461929321, "genuineness": 4.2873719215393065, "identity": 0.5591023281216622, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.044098880141973495, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "config": { "scale": 1.0, "cfg": 1.0 }, "n": 20, "target_z": 1.2098632977825212, "target_raw": 1.1541010707616806, "ce": 6.24063994884491, "genuineness": 4.3350698232650755, "identity": 0.5654936224222183, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.05049017444252968, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" } ], "development_standard": [ { "scale": 0.0, "n": 20, "target_z": 1.0951171675342077, "target_raw": 1.0908753991127014, "ce": 6.2072807312011715, "genuineness": 4.368960380554199, "identity": 0.5150034479796887, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.0, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 0.5, "n": 20, "target_z": 0.9869140663241359, "target_raw": 1.0312549680471421, "ce": 6.179282832145691, "genuineness": 4.3768054962158205, "identity": 0.5431524679064751, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.028149019926786423, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 0.75, "n": 20, "target_z": 1.0262695164288325, "target_raw": 1.052940011024475, "ce": 6.158036136627198, "genuineness": 4.294275569915771, "identity": 0.5801795639097691, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.06517611593008041, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" }, { "scale": 1.0, "n": 20, "target_z": 0.9674804828950772, "target_raw": 1.0205469697713851, "ce": 6.190204000473022, "genuineness": 4.357663571834564, "identity": 0.5732357978820801, "wer": 0.0, "valid_fraction": 1, "identity_delta": 0.058232349902391435, "identity_pairs": 20, "gates_pass": true, "gate_scope": "development_provisional_20_pairs" } ] }, "heldout": { "standard": { "n": 50, "target_z": 1.1895898349333596, "target_raw": 1.1429302948713302, "ce": 6.2137142848968505, "genuineness": 4.343165912628174, "identity": 0.5770200660824776, "wer": 0.024086698960840838, "valid_fraction": 0.98, "identity_delta": 0.0, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" }, "manual": { "n": 50, "target_z": 2.082812522750439, "target_raw": 1.6351003074645996, "ce": 6.2736020755767825, "genuineness": 4.122876906394959, "identity": 0.5433347749710083, "wer": 0.011742449007895232, "valid_fraction": 0.98, "identity_delta": -0.03368529111146927, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" }, "adapter": { "n": 50, "target_z": 2.256601569806001, "target_raw": 1.7308589112758637, "ce": 6.239766302108765, "genuineness": 4.03871518611908, "identity": 0.5451197719573975, "wer": 0.016878460974570814, "valid_fraction": 1, "identity_delta": -0.03190029412508011, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" }, "matched_no_adapter": { "n": 50, "target_z": 2.2890624747480266, "target_raw": 1.7487450265884399, "ce": 6.223245525360108, "genuineness": 4.057104024887085, "identity": 0.5571359089016914, "wer": 0.007527811623921464, "valid_fraction": 1, "identity_delta": -0.019884157180786132, "identity_pairs": 50, "gates_pass": true, "gate_scope": "heldout_50_reference_pairs" } }, "adapter_vs_matched_cfg": { "n": 50, "delta_z": -0.032460904942025526, "ci95_z": [ -0.1382031033735609, 0.09523443945676088 ], "language_deltas": { "en": 0.012890576025707434, "de": -0.11660152220924247, "fr": 0.21562512879520596, "nl": -0.09062494219427764, "es": -0.18359376512752093 }, "languages_improved": 2, "clear_gain": false, "delta_ce": 0.016520776748657227, "delta_genuineness": -0.018388838768005372, "delta_wer": 0.00935064935064935, "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": 0.17378904705556175, "ci95_z": [ 0.04460938259756497, 0.33832028875655307 ], "language_deltas": { "en": 0.24023428638991245, "de": 0.033398450747144745, "fr": 0.4726562355485694, "nl": 0.08906251369082896, "es": 0.03359374890135321 }, "languages_improved": 5, "clear_gain": true, "delta_ce": -0.033835773468017576, "delta_genuineness": -0.08416172027587891, "delta_wer": 0.005136011966675582, "note": "Exploratory language-cluster bootstrap; five voices, one fixed reference voice per language. No familywise claim across72targets." }, "manual_vs_standard": { "n": 50, "delta_z": 0.8932226878170796, "ci95_z": [ 0.6931250441486979, 1.135683634947141 ], "language_deltas": { "en": 0.5994140839024842, "de": 0.7281250327903804, "fr": 0.81054698735776, "nl": 1.3524414000067284, "es": 0.9755859350280448 }, "languages_improved": 5, "clear_gain": true, "delta_ce": 0.05988779067993164, "delta_genuineness": -0.22028900623321535, "delta_wer": -0.012344249952945606, "note": "Exploratory language-cluster bootstrap; five voices, one fixed reference voice per language. No familywise claim across72targets." }, "route": { "adapter": false, "scale": 0.0, "cfg": 4.0, "reason": "Use manual-only configuration; no clear qualifying incremental adapter gain" }, "reference_only": true, "updated": 1793043299.1747646, "example": { "text": "De weg gaat linksaf na de brug. Loop verder... tot aan het kleine houten hek. Ik wacht aan de andere kant.", "sentences": [ "De weg gaat linksaf na de brug.", "Loop verder... tot aan het kleine houten hek.", "Ik wacht aan de andere kant." ], "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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.", "language": "nl", "seed": 1702, "scale": 0.0, "cfg": 4.0, "frames": 128, "reference": true } }, "rows": [ { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "0000fd8a782accd9f828407b2142e9915e7d861794108bd6a4582d6ba476a6a2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "340e09d14cbd8c1f3fe7ae491263a1b6ab1fbb94814db11c8ed41f1763eff96b", "generation_seconds_batch": 20.34294045297429, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1934354305267334, "target_z": 1.2812498952060003, "ce": 6.109982967376709, "genuineness": 4.293961524963379, "identity": 0.7032514810562134, "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_NEWS_reg/0000fd8a782accd9f828407b2142e9915e7d861794108bd6a4582d6ba476a6a2.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": "0030343ac333e0814aacd5fdfa4ffad061b10f2b2c53070d835483244ffe10dd", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "aced2b4f35cbc2d4e629f540add93df2f5ed14c5ab2c6573d60fac6848883fd2", "generation_seconds_batch": 14.240347187966108, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.908019781112671, "target_z": 2.578124832667646, "ce": 5.982900142669678, "genuineness": 3.9550557136535645, "identity": 0.45217692852020264, "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_NEWS_reg/0030343ac333e0814aacd5fdfa4ffad061b10f2b2c53070d835483244ffe10dd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "00352d0c0afc1219042ca641309f81e9a3937c4be299064d13fb190df59f8bdf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "36ce433312359b7b82051702f3186c5785646816a00c3605af2bb0ee6aa909b0", "generation_seconds_batch": 15.188897399930283, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.451719045639038, "target_z": 1.7500000811308385, "ce": 6.213237762451172, "genuineness": 4.419482707977295, "identity": 0.6103384494781494, "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_NEWS_reg/00352d0c0afc1219042ca641309f81e9a3937c4be299064d13fb190df59f8bdf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "004ff8300d68eee8d896ddc77688705f62ba5abfc19f3ca98db5889b1a21b911", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "351700c72fe5d5052644d6edc54fa56bc3d940cc72caf49ff3881b391ecbbf0f", "generation_seconds_batch": 20.867851946968585, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5292041301727295, "target_z": 1.8906251369082898, "ce": 5.903473377227783, "genuineness": 4.231201171875, "identity": 0.5236647129058838, "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_NEWS_reg/004ff8300d68eee8d896ddc77688705f62ba5abfc19f3ca98db5889b1a21b911.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": "00526f756bdf4a1519d5346cbe6e55bc03be517834e7f7739f65b27daa73aeaa", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9bd4da6c9f16526a7141fa18b1d5c988fdc564552f84322f2ff7ea1c99dd9e54", "generation_seconds_batch": 13.988149769953452, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 6.1894097328186035, "genuineness": 4.231201171875, "identity": 0.7824846506118774, "hypothesis": "De weg gaat linksaf na de brug. Loop verder tot aan het kleine houten hek. Ik wacht aan de andere kant.", "wer": 0.0, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/00526f756bdf4a1519d5346cbe6e55bc03be517834e7f7739f65b27daa73aeaa.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "00ab088ee186d7a66d4b6bc6d7121619fd1fef872e5c9584009df62b724058cb", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "664a538db2ea014894bea8a0bd3149de18c3797788dab2207e64ee53b0027797", "generation_seconds_batch": 14.041933526983485, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5119850635528564, "target_z": 1.8593748360480973, "ce": 5.816103935241699, "genuineness": 3.9173994064331055, "identity": 0.5261260271072388, "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_NEWS_reg/00ab088ee186d7a66d4b6bc6d7121619fd1fef872e5c9584009df62b724058cb.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "00db7896667dbdf06abf8b93ffe2033ffc7697f861ff23150369eabb648f5dca", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3b61e98ff514063c1f246930861eb4dbf2e700938a30fa8ee2ad8f1952498a18", "generation_seconds_batch": 14.321042104973458, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0384653806686401, "target_z": 1.0, "ce": 6.197352409362793, "genuineness": 4.607763767242432, "identity": 0.6771131753921509, "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_NEWS_reg/00db7896667dbdf06abf8b93ffe2033ffc7697f861ff23150369eabb648f5dca.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": "010a3bba9a0fc48b6605dd47d23439b538517ed34db04b501a7299730f2db563", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The road turns left behind the bridge.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c99aad8bb0820d7abf178181cfe1c42f4df8c23ba45e65d423e2e374ad8e9394", "generation_seconds_batch": 14.21999103098642, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2364827394485474, "target_z": 1.3593749983097743, "ce": 6.435632705688477, "genuineness": 3.8295347690582275, "identity": 0.32290899753570557, "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_NEWS_reg/010a3bba9a0fc48b6605dd47d23439b538517ed34db04b501a7299730f2db563.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "010d221789304377baa5220879c695ba676f12ff342b6399f4be8b1ab8a20146", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "90fd353076e5d9fdeb555a074e98599df3db17c1741df7983d5b2a38d17f9e81", "generation_seconds_batch": 18.867784505011514, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4043669700622559, "target_z": 1.6640624028120166, "ce": 6.086154937744141, "genuineness": 4.243753433227539, "identity": 0.7498027086257935, "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_NEWS_reg/010d221789304377baa5220879c695ba676f12ff342b6399f4be8b1ab8a20146.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "1b597d7291e2ad6846c571f86cacd5efd73e7cd6392d6e46e6bd3912ef4ad9a7", "generation_seconds_batch": 13.90252202108968, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.537813425064087, "target_z": 1.9062498546405813, "ce": 6.109982967376709, "genuineness": 4.406930446624756, "identity": 0.7458639144897461, "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_NEWS_reg/01f5d5f22d9b51f4e674eec92be80d0710e62f5704093e3409bc811236eec3e5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "020bfeb7ca55103458576e998e12595cdfa95989e9d9de723ffad8d849afdefa", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "feb06209f29060db7a910ff5248b8faf0b228d61fd81510e5d3df62918373974", "generation_seconds_batch": 14.941111410036683, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5248992443084717, "target_z": 1.8828123453443393, "ce": 6.197352409362793, "genuineness": 4.243753433227539, "identity": 0.40369486808776855, "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_NEWS_reg/020bfeb7ca55103458576e998e12595cdfa95989e9d9de723ffad8d849afdefa.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "0219f853ffa16ae9fdf0e9806bae7e9a67e5186783bdfe1035527c559eaff114", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5d3bcb713c9fbc6bf39bc80dfc09f7a1cb1c02d6a35ad2e8fe0d19ee69ef4090", "generation_seconds_batch": 18.66631546197459, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.468937873840332, "target_z": 1.7812499492932259, "ce": 6.1894097328186035, "genuineness": 4.306513786315918, "identity": 0.5135461688041687, "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_NEWS_reg/0219f853ffa16ae9fdf0e9806bae7e9a67e5186783bdfe1035527c559eaff114.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "02244b5fd6ce28ed3e0cf90f14dd6af67adc7a4d1bd2e09c7ff4fa02b9b6c938", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2a08a060083e609807a258b5b46d36fc4aa93f418954db9af1aca09dca483ba4", "generation_seconds_batch": 20.338860796997324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.372091293334961, "genuineness": 4.086852550506592, "identity": 0.4630678594112396, "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_NEWS_reg/02244b5fd6ce28ed3e0cf90f14dd6af67adc7a4d1bd2e09c7ff4fa02b9b6c938.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "028b7d391ed2d04b322c56fbb8d9dbeef3b40bc75f86088cebee312dd1dcb0e9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8676405b4ce3a901182f230344b011cfce45bca4221e690a493bbdb6e5362429", "generation_seconds_batch": 20.51661712594796, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7702686786651611, "target_z": 2.3281250219729355, "ce": 6.173524379730225, "genuineness": 4.168440818786621, "identity": 0.6918300986289978, "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_NEWS_reg/028b7d391ed2d04b322c56fbb8d9dbeef3b40bc75f86088cebee312dd1dcb0e9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "02960d7ab7d00e61a86da371f245c05e9107caaf0f09d04ab24ee8b6cd900060", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d7c500fcca55bdb5726f4e6a4af1ac5227953ff1b133294b2f7ac72c9d047b30", "generation_seconds_batch": 19.018228149972856, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8469051122665405, "target_z": 0.6523437563383467, "ce": 6.435632705688477, "genuineness": 4.20609712600708, "identity": 0.5753371715545654, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/02960d7ab7d00e61a86da371f245c05e9107caaf0f09d04ab24ee8b6cd900060.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "03b90c33a5c94b7cac6de3763f617a459150880a09608ddaca9a6110ef124601", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "365b8e3a82787e30ffa175d3ec547f7739e5d15decaad5503f5ce9ca07f4f4df", "generation_seconds_batch": 14.319818526040763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9166293144226074, "target_z": 2.593749983097742, "ce": 6.372091293334961, "genuineness": 3.8797430992126465, "identity": 0.583713173866272, "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_NEWS_reg/03b90c33a5c94b7cac6de3763f617a459150880a09608ddaca9a6110ef124601.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "03c7ad1c835e44071dab899a6ba3cdec930f9060e00b4a254c8abbaa9184feca", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f75608ccb5aae6a4edc9525966539e61405539cac02f6f8952bf9ebf1d6290a6", "generation_seconds_batch": 20.65841965796426, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.606689214706421, "target_z": 2.031250192685741, "ce": 6.3641486167907715, "genuineness": 4.105680465698242, "identity": 0.5906503200531006, "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_NEWS_reg/03c7ad1c835e44071dab899a6ba3cdec930f9060e00b4a254c8abbaa9184feca.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": "043e3ebacab5f3c28fac819e96276bd870e1801cf62fa72e0822008c415cb121", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Escribí el número en el sobre.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "52ca6024590d314a52ecae578fbde78dbffa6e6e87b76374500a516069d56c15", "generation_seconds_batch": 19.02570185402874, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.572251319885254, "target_z": 1.9687500236631612, "ce": 6.229123115539551, "genuineness": 4.218649387359619, "identity": 0.5350341796875, "hypothesis": "Escribí el número en el sobre, léelelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/043e3ebacab5f3c28fac819e96276bd870e1801cf62fa72e0822008c415cb121.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": "047825685956e10266694351b6aef2fc6978fe8a75cda0985dbeffc63cba70a9", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The road turns left behind the bridge.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0a12cbc4355d5c271d3809596cb3198cb86e376414dc780ca343b073e17e34ba", "generation_seconds_batch": 21.403163607930765, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2709205150604248, "target_z": 1.4218749509834518, "ce": 6.435632705688477, "genuineness": 3.6851859092712402, "identity": 0.03558605909347534, "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_NEWS_reg/047825685956e10266694351b6aef2fc6978fe8a75cda0985dbeffc63cba70a9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "04b9ec4c97ca7dace40e7389646eda97922c0afb5cdf2f99923a7247e55d8c01", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "72980bf58eb1435bc927c1d7453f423e9bee33f9a9cadbe745b15cefc405542a", "generation_seconds_batch": 19.014809328014962, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.417281150817871, "target_z": 1.6874999121082583, "ce": 6.078212261199951, "genuineness": 4.243753433227539, "identity": 0.6707266569137573, "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_NEWS_reg/04b9ec4c97ca7dace40e7389646eda97922c0afb5cdf2f99923a7247e55d8c01.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "04f6f885100533599d314a72b815db24aa65d8beb03355a94674be7085f12cc3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9efd11dc998dffde8178dccbd95e99e72956b15156497db0c3cf9a128a62c5aa", "generation_seconds_batch": 19.951011589029804, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4947662353515625, "target_z": 1.8281249678857099, "ce": 6.070269584655762, "genuineness": 4.15588903427124, "identity": 0.6274905204772949, "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_NEWS_reg/04f6f885100533599d314a72b815db24aa65d8beb03355a94674be7085f12cc3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "055250a96124d471972b6b7f517ff2dc33aca5e1b172df94fa9cd2763ece3f6d", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9c4e75bba3e3b58f56e018df94c90b70e894dd8d0cebdc386a418d07762cbf2d", "generation_seconds_batch": 15.893671412952244, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6927835941314697, "target_z": 2.187499966195484, "ce": 6.181467056274414, "genuineness": 4.030367851257324, "identity": 0.765418529510498, "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_NEWS_reg/055250a96124d471972b6b7f517ff2dc33aca5e1b172df94fa9cd2763ece3f6d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "057e97faa4497ac509323ee2d4d252124aa7f62965f555e039c68d264d5ed842", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b74edc33bf8f22a3b567505e23253913a2ba0b90a810dafd20e4dbfaf5d501fb", "generation_seconds_batch": 21.057972837006673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.133169412612915, "target_z": 1.1718751402887415, "ce": 6.09409761428833, "genuineness": 4.2374773025512695, "identity": 0.7081354856491089, "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_NEWS_reg/057e97faa4497ac509323ee2d4d252124aa7f62965f555e039c68d264d5ed842.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "458b1b1a0c0d47b43669ae1c0739403af04289505e5f7c4a40824e4537768fdd", "generation_seconds_batch": 15.010534388013184, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5262030959129333, "target_z": 0.0703125278887257, "ce": 6.284721851348877, "genuineness": 4.494794845581055, "identity": 0.6367429494857788, "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_NEWS_reg/059867605942d1f4c46c2c44d936e08b7fbe89dd4feef4c05500a7ed426c27a3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "07889d99e03888dcef3d3e51f777a4a0956edfdc4a2970e1b966fc01ea97e3f6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "12d4543161afd9ea53f0dc74deef37565d16d8ffdd4f963d543aaf42b41e6977", "generation_seconds_batch": 20.610676417010836, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8167720437049866, "target_z": 0.5976562707052661, "ce": 6.221180438995361, "genuineness": 4.080576419830322, "identity": 0.2541748881340027, "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_NEWS_reg/07889d99e03888dcef3d3e51f777a4a0956edfdc4a2970e1b966fc01ea97e3f6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "078edf9d923c5bb5ebb8984f8a2db5fab3d3927c0f4ddb0cde5c33dc380bf7bf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "971744b23a57649033dd0b7d30b8893d4e9e59d4421e5b4b925fb793e966aedb", "generation_seconds_batch": 19.179725417052396, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.85636305809021, "target_z": 2.4843747954826783, "ce": 6.173524379730225, "genuineness": 4.18099308013916, "identity": 0.7559947967529297, "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_NEWS_reg/078edf9d923c5bb5ebb8984f8a2db5fab3d3927c0f4ddb0cde5c33dc380bf7bf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "07c8b4d9ac3ee1c1f5ce2b13e54325911f59fd33b97e7e4bbe79eb1f81f1da27", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4c4c98993bf30616320d152697e5b2bea82c6e1264ccd698548c73e198ccb6b9", "generation_seconds_batch": 19.25732411595527, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4947662353515625, "target_z": 1.8281249678857099, "ce": 6.443574905395508, "genuineness": 4.262581825256348, "identity": 0.6115320920944214, "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_NEWS_reg/07c8b4d9ac3ee1c1f5ce2b13e54325911f59fd33b97e7e4bbe79eb1f81f1da27.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": "082d223e171819464a59b3780101e5ea3d68642d1d5ba26e3ee599869e775e82", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "ab734721aebf32d16efdcf564063767c1729dedccd8d0ba38a52a93e87a8ecd2", "generation_seconds_batch": 14.911016997066326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1848260164260864, "target_z": 1.2656249611248065, "ce": 6.133810997009277, "genuineness": 4.369274139404297, "identity": 0.6245514154434204, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/082d223e171819464a59b3780101e5ea3d68642d1d5ba26e3ee599869e775e82.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": "08965d1ee5268f7df33bc064756dbafc2bb800424c88491f380eda17ae0c5277", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "09a92bcbfecab47d36f99549c43db669b39e8c67a5d4c8eade505ef9d5d74638", "generation_seconds_batch": 13.942868506070226, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.329882860183716, "target_z": 3.343749847879678, "ce": 6.38797664642334, "genuineness": 4.268857479095459, "identity": 0.6351406574249268, "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_NEWS_reg/08965d1ee5268f7df33bc064756dbafc2bb800424c88491f380eda17ae0c5277.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "08e9c3eb11a3a59263f45b16f33daeafc326da6c68872c11953e2965c59c99c0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7d0657b70ff2521b1a22d578f6a8203d1a639483b5fea63fcf67fa3144e3ae84", "generation_seconds_batch": 15.392544595059007, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.047074794769287, "target_z": 1.0156249340811938, "ce": 6.197352409362793, "genuineness": 3.835810661315918, "identity": 0.2238512933254242, "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_NEWS_reg/08e9c3eb11a3a59263f45b16f33daeafc326da6c68872c11953e2965c59c99c0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "08f80bbc79809613a514db16afeef3f20dfd2195b79c7d81270b04f728303973", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "78e4f8135325b3ceb3bbcb26b87a905cf5c7ac9adb6f74ff28de37c5e8532edc", "generation_seconds_batch": 15.922158724046312, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0599889755249023, "target_z": 1.0390624433774358, "ce": 6.117925643920898, "genuineness": 4.532451152801514, "identity": 0.6479377150535583, "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_NEWS_reg/08f80bbc79809613a514db16afeef3f20dfd2195b79c7d81270b04f728303973.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": "0918346b71b5eb6690eaa6ec60e8282c48c43494f0b8c79d241e3a20c5c13af4", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c418011477f2fcfdab7c1df79d8e402ff39a370ffe2ed78b5a084e147e193fe8", "generation_seconds_batch": 14.929697441053577, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4775474071502686, "target_z": 1.7968750997233223, "ce": 5.967014789581299, "genuineness": 4.13078498840332, "identity": 0.6532991528511047, "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_NEWS_reg/0918346b71b5eb6690eaa6ec60e8282c48c43494f0b8c79d241e3a20c5c13af4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "091d294b211436bdd5878824ac726acfa3541551746c2dd44d348a416f1555c8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "727220a9e9190d027934f9f8a5548b6faa82af885a2540a97c360dafc91091a6", "generation_seconds_batch": 20.48795889201574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.318272590637207, "target_z": 1.5078126293022738, "ce": 6.260893821716309, "genuineness": 4.369274139404297, "identity": 0.5496019124984741, "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_NEWS_reg/091d294b211436bdd5878824ac726acfa3541551746c2dd44d348a416f1555c8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "0934dcd6b59bea309d1e703e81bc4de54a1996d8029fedd077e9deb3faf4e1ff", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "239c1df45a2581947bfcad5ea8a92e3fce0dfecf6b7f5477decf353e7a82f559", "generation_seconds_batch": 18.607872123015113, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7694200277328491, "target_z": 0.5117187005608953, "ce": 6.3641486167907715, "genuineness": 4.281409740447998, "identity": 0.4579177498817444, "hypothesis": "I found Little Box. It was beside the old window. We can look inside together.", "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_NEWS_reg/0934dcd6b59bea309d1e703e81bc4de54a1996d8029fedd077e9deb3faf4e1ff.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "0948a1cf99fd47f78bc2cb2bb12a81076c916ea819c2c3eb5afc1ec881e77801", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d085d0f51b3d94bbcac3236e25a71bae260e9d31918623443c0da0efc8eec618", "generation_seconds_batch": 20.607717695995234, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4474142789840698, "target_z": 1.7421875059157903, "ce": 6.340320587158203, "genuineness": 4.381826400756836, "identity": 0.2870112955570221, "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_NEWS_reg/0948a1cf99fd47f78bc2cb2bb12a81076c916ea819c2c3eb5afc1ec881e77801.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "09befbfeb6325749e90486b08b5e9f7984ade4c4f974ab16d3cd53caa5851116", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.75, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "afcb03d169f06cfbcb9bb4a09fc6ddfcd81199cadcae7a5ec3e44b037d707e8f", "generation_seconds_batch": 19.32386030408088, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2063496112823486, "target_z": 1.3046874045022423, "ce": 6.149696350097656, "genuineness": 4.250029563903809, "identity": 0.4409947097301483, "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_NEWS_reg/09befbfeb6325749e90486b08b5e9f7984ade4c4f974ab16d3cd53caa5851116.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": "0a087c2e9eaa7d3b203a6eba86d01002894a79a0770540f3a55083e16c39b649", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f666e079cab213d7d5058905feb37314cdc3ad3aa4fd150f9d117e056784625a", "generation_seconds_batch": 14.068191012949683, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.503375768661499, "target_z": 1.843750118315806, "ce": 6.213237762451172, "genuineness": 4.419482707977295, "identity": 0.6317901611328125, "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_NEWS_reg/0a087c2e9eaa7d3b203a6eba86d01002894a79a0770540f3a55083e16c39b649.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "0a0e84d2e39bfd5aab9e1b0e60929973ecbe5bdf976248de8e9dad42dd211e1d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "09907a377754f8ca7e8ca82ab477cd55a3cd2caaf929243f555583210c544f63", "generation_seconds_batch": 14.887773491092958, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7629629373550415, "target_z": 0.4999999459127744, "ce": 6.229123115539551, "genuineness": 4.055472373962402, "identity": 0.1425212323665619, "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_NEWS_reg/0a0e84d2e39bfd5aab9e1b0e60929973ecbe5bdf976248de8e9dad42dd211e1d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "0abcd816a7e2958d529cde0b58a2c92df109073bafcc4608be79fca39e19602f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6d72745162c95387f947f33daf62ea316619fc6790197664fcdd001cb45ff07d", "generation_seconds_batch": 15.165268667973578, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4861568212509155, "target_z": 1.812500033804516, "ce": 6.125868320465088, "genuineness": 4.231201171875, "identity": 0.5981196165084839, "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_NEWS_reg/0abcd816a7e2958d529cde0b58a2c92df109073bafcc4608be79fca39e19602f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "0b237093007f3d322f5ce72774bffb78c962e3a47b8e54e8a6a7f2abec247a9d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ed624d51bdd42fab7dae4334958c5f0f41fce561142d4368074fe837d7873893", "generation_seconds_batch": 21.487543350085616, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4818520545959473, "target_z": 1.8046874585894679, "ce": 5.919358730316162, "genuineness": 4.262581825256348, "identity": 0.5991185903549194, "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_NEWS_reg/0b237093007f3d322f5ce72774bffb78c962e3a47b8e54e8a6a7f2abec247a9d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "0b4ea4f0427a5d45f9c5d06a120ef7ac038f2940d01cefccc1d763cd232acbb6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ee6007e9175f45e2f82983c50ce035f252d502f90658c144722ddad01410763d", "generation_seconds_batch": 15.394675730960444, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5464229583740234, "target_z": 1.9218750050706774, "ce": 5.8875885009765625, "genuineness": 4.2374773025512695, "identity": 0.4784421920776367, "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_NEWS_reg/0b4ea4f0427a5d45f9c5d06a120ef7ac038f2940d01cefccc1d763cd232acbb6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "0bd4fb62d96dcac1e722cfa7ce17590754b03083851a344677b27b2c13bff8d7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "451f372f14a43e55a40b94921a9535bf052d959ebacf616dfb87c87d9e8753f3", "generation_seconds_batch": 21.55270903604105, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1589977741241455, "target_z": 1.2187501588812253, "ce": 6.3641486167907715, "genuineness": 4.532451152801514, "identity": 0.5341721773147583, "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_NEWS_reg/0bd4fb62d96dcac1e722cfa7ce17590754b03083851a344677b27b2c13bff8d7.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "d6f5dc386fbef8f625849fd89b664a4ebd2e3805ccbb41dfc800d5e306acd23e", "generation_seconds_batch": 14.746553199947812, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3742339611053467, "target_z": 1.609375025353387, "ce": 6.252951145172119, "genuineness": 4.570107460021973, "identity": 0.8007564544677734, "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_NEWS_reg/0cb70e86df8754d3bba44fe7f798aac7d0ffe2ca68fc75b92ff1a1ed29986ba7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "0d82c965a61af9030aa04a58f75ac28f48ef98d900e3bb410f0e7376e7c739bf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "fb623c9c25328b252b5868fc9e78eff9604da091e8b163479096eb1562fc6c33", "generation_seconds_batch": 18.68280913599301, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3053584098815918, "target_z": 1.4843751200060318, "ce": 6.332377910614014, "genuineness": 4.093128204345703, "identity": 0.2946954369544983, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/0d82c965a61af9030aa04a58f75ac28f48ef98d900e3bb410f0e7376e7c739bf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "0da064e16f2b593d04e95b5977e727a4bf8600165939aaceb4701b2d13b3af11", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b63a62ed03118bcd71759258036e60be356c55a01c6a9ccbc2237fec5baf49aa", "generation_seconds_batch": 16.530102399061434, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9825039505958557, "target_z": 0.8984374957744355, "ce": 6.181467056274414, "genuineness": 4.457139015197754, "identity": 0.7035318613052368, "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_NEWS_reg/0da064e16f2b593d04e95b5977e727a4bf8600165939aaceb4701b2d13b3af11.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "0db2bd3fed919f5af9df8f6b52d14c67d4ebff538806f1f51a92a207068667b5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3b5115975b2423ac9d0cb0a5640fbd999891d01ad21a4449928ec91ae39d1b3f", "generation_seconds_batch": 18.87896600307431, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6467353701591492, "target_z": 0.2890624704210485, "ce": 6.268836498260498, "genuineness": 4.7458367347717285, "identity": 0.7029420733451843, "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_NEWS_reg/0db2bd3fed919f5af9df8f6b52d14c67d4ebff538806f1f51a92a207068667b5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "0dc32cba06fd74e87a9694c1431d4cdda64a21948e583f9ae83fb4576ef821ce", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "53104c54bd9b9a7a27f88f1aabb4046fbb6ac0929c488d0077988ff622fd3c7e", "generation_seconds_batch": 20.37464256596286, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4388048648834229, "target_z": 1.7265625718345965, "ce": 6.395919322967529, "genuineness": 4.419482707977295, "identity": 0.427864670753479, "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_NEWS_reg/0dc32cba06fd74e87a9694c1431d4cdda64a21948e583f9ae83fb4576ef821ce.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "0e7c852f67885a2b78141435f18a5791534e1a4c8ad0ad877bb6e00bbd821ccf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d2f1a545ad0f853e8f144dea726cac81c06af4082097f2818ff2c486dec2dd4f", "generation_seconds_batch": 14.772921173018403, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1589977741241455, "target_z": 1.2187501588812253, "ce": 6.435632705688477, "genuineness": 4.381826400756836, "identity": 0.2817440330982208, "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_NEWS_reg/0e7c852f67885a2b78141435f18a5791534e1a4c8ad0ad877bb6e00bbd821ccf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "0e88c0521686ba3b172c44574d62628a9291925556d3d4be0e9cab2844d945ae", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c2d3ca3331044485855668eb1a042fa6526406c421a44754f8a0d20fe8511122", "generation_seconds_batch": 14.405578007921576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5765559673309326, "target_z": 1.9765623825293068, "ce": 6.09409761428833, "genuineness": 4.15588903427124, "identity": 0.7010437250137329, "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_NEWS_reg/0e88c0521686ba3b172c44574d62628a9291925556d3d4be0e9cab2844d945ae.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "0eb834880ec671c5514d486ac133de965d907cd938e5bffe66608135818f323a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "e6f6712fddc028cb04c8ed9a33342ade09b050ce3877d851806bd91161905333", "generation_seconds_batch": 20.168585931998678, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4086717367172241, "target_z": 1.6718749780270645, "ce": 6.149696350097656, "genuineness": 4.331617832183838, "identity": 0.7016699910163879, "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_NEWS_reg/0eb834880ec671c5514d486ac133de965d907cd938e5bffe66608135818f323a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "0efb1027298b1b64e87e2677e4d72803b96104070afd89e3f986fd601d1c6f39", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b856e24a8c09229fd00a1b3c48967ddc7f36d8c6a68e5bc796db34b57836a680", "generation_seconds_batch": 19.76222982001491, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2696168422698975, "target_z": 3.234375092962419, "ce": 6.324435234069824, "genuineness": 3.8420867919921875, "identity": 0.5988026857376099, "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_NEWS_reg/0efb1027298b1b64e87e2677e4d72803b96104070afd89e3f986fd601d1c6f39.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "0f277a62fe88d608f64de161780773c78a6d8601bcd42b399b2230951a96dc3e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "976c278117b4d912020435e2076aa0ea5ef2cfc4fd693e90a9d6419673952098", "generation_seconds_batch": 20.212513353093527, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6927835941314697, "target_z": 2.187499966195484, "ce": 6.260893821716309, "genuineness": 4.231201171875, "identity": 0.6040621399879456, "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_NEWS_reg/0f277a62fe88d608f64de161780773c78a6d8601bcd42b399b2230951a96dc3e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "101e53033d81cf9e48160da6109d17847582b849533c2dd2bb27fc499f2e1852", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c7c494db8bec158cd47641b73753a58a552a4d45d806d214f4401db3e05381e2", "generation_seconds_batch": 20.237919704057276, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7960970401763916, "target_z": 2.375000040565419, "ce": 5.998785495758057, "genuineness": 3.9425034523010254, "identity": 0.49721091985702515, "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_NEWS_reg/101e53033d81cf9e48160da6109d17847582b849533c2dd2bb27fc499f2e1852.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": "10de4bf95d88be808745024af91ed90fc3cc4e814ec44ede21a3369ede50a6b5", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Someone moved the little blue chair.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "aead059ae354c966c50bdf05f48ac61c1af930b7564cd256c1c1da21e47432b1", "generation_seconds_batch": 20.91047477896791, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.787487506866455, "target_z": 2.359374890135323, "ce": 6.538887023925781, "genuineness": 4.18099308013916, "identity": 0.30983784794807434, "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_NEWS_reg/10de4bf95d88be808745024af91ed90fc3cc4e814ec44ede21a3369ede50a6b5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "1131a733033e08151abc166d15d64faf063081d4e82fc6cf49b96d15bff09aff", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "18adb81c49c28ac7f153850f0aad52ceaf69a32b6e8cfefd4d9453423d75d061", "generation_seconds_batch": 20.986776006990112, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1146466732025146, "target_z": 2.953124981407516, "ce": 6.1894097328186035, "genuineness": 4.174716949462891, "identity": 0.7048642635345459, "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_NEWS_reg/1131a733033e08151abc166d15d64faf063081d4e82fc6cf49b96d15bff09aff.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": "1132e40d80c8cac9c32546eec40fd322cf53e80e245b13f2a8b87d72f8de1064", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d2f0d6b1be1abeee4660495c0b691fad8cfc9218c7500d587fa3598c3b5787c2", "generation_seconds_batch": 19.83442033000756, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.261007308959961, "target_z": 3.2187499425323227, "ce": 6.213237762451172, "genuineness": 4.105680465698242, "identity": 0.5950295329093933, "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_NEWS_reg/1132e40d80c8cac9c32546eec40fd322cf53e80e245b13f2a8b87d72f8de1064.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "127c62616a3428d9727838f9271f0e8d8b825dbcf8394edc2ef4bf6a21c895f6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b8d641d520fa7656ce34dbe68e73c066162095f6c75519dbdf01247dd1e368ec", "generation_seconds_batch": 20.3575299680233, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9588279724121094, "target_z": 0.8554687647894758, "ce": 6.181467056274414, "genuineness": 4.532451152801514, "identity": 0.689797043800354, "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_NEWS_reg/127c62616a3428d9727838f9271f0e8d8b825dbcf8394edc2ef4bf6a21c895f6.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "947ce07b08121db088d48ce1a9583a02441216ec6681c49d5d52051e04a3ead5", "generation_seconds_batch": 14.950288043939508, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 6.125868320465088, "genuineness": 4.118232727050781, "identity": 0.6417236328125, "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_NEWS_reg/12a7b850c68b1e935d69d2b179f3c0df54c4775061935760902fd9f6cbfc7541.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": "13138d07e201e9adf92c83bb77765d5fefc252388d58063b59973c4cd02bd43b", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Escribí el número en el sobre.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7f072b6b7596316f9dcf22e8fe4a16f9f6e6ff6bbf8ff5df00071f261869d1ac", "generation_seconds_batch": 19.02570185402874, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 6.2767791748046875, "genuineness": 4.168440818786621, "identity": 0.7437863349914551, "hypothesis": "Escribí el número en el sobre, léelelo bien, y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/13138d07e201e9adf92c83bb77765d5fefc252388d58063b59973c4cd02bd43b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "1318b1ced388f344739e990facc3eb8db21c1a9a2d057630716f5dc880f07fce", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8a085b569990e9baff00354134f221c03c6b497f4e94c470814fd329ca61cd93", "generation_seconds_batch": 19.107333330088295, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5421183109283447, "target_z": 1.9140626462045318, "ce": 6.23706579208374, "genuineness": 4.400654315948486, "identity": 0.6832753419876099, "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_NEWS_reg/1318b1ced388f344739e990facc3eb8db21c1a9a2d057630716f5dc880f07fce.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "1378d01fdab0d0e9b656727ed86cd26ffd0df95f6bb37ce6c40758c36e80f818", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a24698ebdda0e3ef01b924c8f2e1fd4dec37190705f7b9131cb18cdb1a189f3d", "generation_seconds_batch": 21.14479663304519, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4990708827972412, "target_z": 1.8359373267518555, "ce": 6.149696350097656, "genuineness": 4.055472373962402, "identity": 0.4472298324108124, "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_NEWS_reg/1378d01fdab0d0e9b656727ed86cd26ffd0df95f6bb37ce6c40758c36e80f818.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": "13f4abcf93a2ed54dfc9e16cf3cb9b349725436c2bb7bc8af3471c8e00c0d8bd", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c76490871faffc7679267c63b3c436f18868c4e891611e5b348bdf6abb83ca84", "generation_seconds_batch": 19.323444967041723, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3570151329040527, "target_z": 1.5781251571909993, "ce": 6.173524379730225, "genuineness": 4.086852550506592, "identity": 0.655828595161438, "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_NEWS_reg/13f4abcf93a2ed54dfc9e16cf3cb9b349725436c2bb7bc8af3471c8e00c0d8bd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "14b157d326df494bc9694b6da432c6b13d9ddbbe31b4bac9600b9118dedfbf20", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3f59afe137d719db7891b688c9058efee1722ccbfc5aac4a1048e8752febec49", "generation_seconds_batch": 19.771560126915574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2881394624710083, "target_z": 1.4531250354947418, "ce": 6.284721851348877, "genuineness": 4.469690799713135, "identity": 0.7559746503829956, "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_NEWS_reg/14b157d326df494bc9694b6da432c6b13d9ddbbe31b4bac9600b9118dedfbf20.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "14c8f136b5e5b3aa6d1ab52bb1df1d66910f95516372326c36386781aa531cdb", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "81e1069d7f154baf13bc430eea00c9c339de82bc8b52dd92d80bb32f84f6883a", "generation_seconds_batch": 19.377071659080684, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.572251319885254, "target_z": 1.9687500236631612, "ce": 6.260893821716309, "genuineness": 4.406930446624756, "identity": 0.031529128551483154, "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_NEWS_reg/14c8f136b5e5b3aa6d1ab52bb1df1d66910f95516372326c36386781aa531cdb.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": "14da3f771a6e4fac4865750c0eba8ec381355ba2b2575a84a21edcbecd199525", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Your message arrived after I had left.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c06084a5f0bbad9fbea801138485c82d73b0ef6f376f240912ff6846dd0320c1", "generation_seconds_batch": 14.622531900997274, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1417787075042725, "target_z": 1.1874998580210328, "ce": 6.38003396987915, "genuineness": 3.8922953605651855, "identity": 0.22867316007614136, "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_NEWS_reg/14da3f771a6e4fac4865750c0eba8ec381355ba2b2575a84a21edcbecd199525.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "98c8bec82c02e416829ffa8d4a2fbe261ec1b018125b46fca015494afed6694c", "generation_seconds_batch": 14.950263729086146, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0027239322662354, "target_z": 2.7500001893052897, "ce": 6.2767791748046875, "genuineness": 4.218649387359619, "identity": 0.528060257434845, "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_NEWS_reg/15585f3962d31fd159bb39a755593ab24aa0df87e4725d7772f40811258760f5.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": "1566760c06a4f5bc2b5c59b29277a2fe5bfc703565cafb7eac64ff503d76e7d7", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a6b3a1b29a6b7b22e4e87b8295964246838e4a9e17b717fcaff28b946629fee8", "generation_seconds_batch": 20.435180311091244, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.7431366443634033, "target_z": 4.093750145359419, "ce": 6.300607204437256, "genuineness": 3.7416701316833496, "identity": 0.8320294618606567, "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_NEWS_reg/1566760c06a4f5bc2b5c59b29277a2fe5bfc703565cafb7eac64ff503d76e7d7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "157a2eae8606b7212e5d5437c27be2d7976cf4667da79d7110eea9d91f9d5d2f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fa53fef29db8cda48bfe99617e6d5b4153d86366efc5a685b2689d0a863ed071", "generation_seconds_batch": 19.765931225032546, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.090122103691101, "target_z": 1.0937500371849675, "ce": 6.24500846862793, "genuineness": 4.049196243286133, "identity": 0.2977992594242096, "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_NEWS_reg/157a2eae8606b7212e5d5437c27be2d7976cf4667da79d7110eea9d91f9d5d2f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "15eb4ec1dcbb9452a45889f28384acfb32795d19dca6608cd5bc13c0193956bd", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c6d53a36415cfc9de67417e5fb221d4d0b78ac7e4663e133f7d8577458d3cbdc", "generation_seconds_batch": 19.128968948032707, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1116456985473633, "target_z": 1.1328124805624034, "ce": 6.284721851348877, "genuineness": 3.9927117824554443, "identity": 0.34881746768951416, "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_NEWS_reg/15eb4ec1dcbb9452a45889f28384acfb32795d19dca6608cd5bc13c0193956bd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "162faee5ec36629e7fd0be968789bd854afceea0d73d966678266693b3a81279", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6bce90d19f0c48e56ed93724229e75c321c146df2cbbc9c0acd7b423a9622e24", "generation_seconds_batch": 20.937172919046134, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6411268711090088, "target_z": 2.0937499290105164, "ce": 6.252951145172119, "genuineness": 4.2751336097717285, "identity": 0.8169902563095093, "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_NEWS_reg/162faee5ec36629e7fd0be968789bd854afceea0d73d966678266693b3a81279.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "1632cb3ca5f240a4c6797f7c55f5f3596f4024c9ab5100261063edd8990b9b7a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "06a8b8fa3a0e432df70ef41fb9c83d10682a5e9b0866929a59cf59f30a7fc69d", "generation_seconds_batch": 20.34294045297429, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3828433752059937, "target_z": 1.6249999594345808, "ce": 6.149696350097656, "genuineness": 4.293961524963379, "identity": 0.6107781529426575, "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_NEWS_reg/1632cb3ca5f240a4c6797f7c55f5f3596f4024c9ab5100261063edd8990b9b7a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "16556d3378bf524b92be3ece4489597e441a5e7e421ef67c15a14e9095cc1225", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e5ee45883477b5420a98808a1201f3c634474df6f10248e4f3cb12a0b0f5d321", "generation_seconds_batch": 15.422848433954641, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1934354305267334, "target_z": 1.2812498952060003, "ce": 6.1020402908325195, "genuineness": 4.532451152801514, "identity": 0.6478886008262634, "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_NEWS_reg/16556d3378bf524b92be3ece4489597e441a5e7e421ef67c15a14e9095cc1225.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "1659d6bde4aff34b40d2234dcffa48c2c2816de76ecc1d2ec687d183e7567740", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "25ecbbc23fd2c6b6956f6104975c946cc6bf108e7d1e7c8aa73da8172541abe5", "generation_seconds_batch": 19.10297372599598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0556843280792236, "target_z": 1.03125008451129, "ce": 6.1020402908325195, "genuineness": 4.444586753845215, "identity": 0.6880927085876465, "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_NEWS_reg/1659d6bde4aff34b40d2234dcffa48c2c2816de76ecc1d2ec687d183e7567740.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "1695dd6ad6a5f64f9daf095d83623fc59f291c3783b60c30dd7f5cc2b5d80fcf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0ef8078c6efae9ddd86386920f96544cafcd76897084c0a421b10eac4c0813b3", "generation_seconds_batch": 19.014809328014962, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0729031562805176, "target_z": 1.0624999526736776, "ce": 6.054384231567383, "genuineness": 4.394378662109375, "identity": 0.6949765086174011, "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_NEWS_reg/1695dd6ad6a5f64f9daf095d83623fc59f291c3783b60c30dd7f5cc2b5d80fcf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "16dd08605019039043c03972f41d8e3075cdef4e86fa8e05e32254e15b9ec9f4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "085d6552396000595c6c59a2595827a319f0fb933d9c24f788f9ac7a7704a5fe", "generation_seconds_batch": 20.549167230026796, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 5.9431867599487305, "genuineness": 4.118232727050781, "identity": 0.7052654027938843, "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_NEWS_reg/16dd08605019039043c03972f41d8e3075cdef4e86fa8e05e32254e15b9ec9f4.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": "16e789b026a43bb79b28d4421d7ca8a52de90afbd22a04d6f15d8738b33674dc", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The road turns left behind the bridge.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "72439c3a32880dac2da1a6d4e8c03a096b0993e3b2de201c0285e544af2748f9", "generation_seconds_batch": 15.90978264901787, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.41180419921875, "genuineness": 3.8671910762786865, "identity": 0.24328097701072693, "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_NEWS_reg/16e789b026a43bb79b28d4421d7ca8a52de90afbd22a04d6f15d8738b33674dc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "1757442db091ba86b4eb41bfb726325a125c02295fdf48b4b229b3634b2bbdd4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "8ba9fa8c1e4d3bf860f915836e47a836c4c3f737b8a209abc4f6488d7176838a", "generation_seconds_batch": 16.003847575979307, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4086717367172241, "target_z": 1.6718749780270645, "ce": 6.09409761428833, "genuineness": 4.2751336097717285, "identity": 0.6708236932754517, "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_NEWS_reg/1757442db091ba86b4eb41bfb726325a125c02295fdf48b4b229b3634b2bbdd4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "175e9877a87386ca0faba72e4ec6f8a654ff1b9169e75735771264cf681cbd5f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "333b68ca4521f8980c92e5bf882e0ccd4e23fe728f150bdd450083d9c2bad142", "generation_seconds_batch": 19.12058468698524, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0685985088348389, "target_z": 1.054687593807532, "ce": 6.467403411865234, "genuineness": 4.293961524963379, "identity": 0.6122453212738037, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/175e9877a87386ca0faba72e4ec6f8a654ff1b9169e75735771264cf681cbd5f.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": "18948ae5ff19785b6d9a5f7e3347b939bf280dd19fe890015d9bb8bce63ec4bd", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d78089c629c3b033e071e325397843c7b8f9a5eab0e8ddfeff559fc34af1a9fa", "generation_seconds_batch": 19.612351148040034, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.261007308959961, "target_z": 3.2187499425323227, "ce": 6.3641486167907715, "genuineness": 3.998987913131714, "identity": 0.4866749346256256, "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_NEWS_reg/18948ae5ff19785b6d9a5f7e3347b939bf280dd19fe890015d9bb8bce63ec4bd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "18a07cf363bdf1047234f78be9c6e653709fd6b0be4c8c90fcbf383a6d89b264", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7770e0e90d093802bf777d90c7e71261d50f32386a52b7ab9fb221654bdd141d", "generation_seconds_batch": 18.77553105202969, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2967488765716553, "target_z": 1.4687499695759356, "ce": 6.022613525390625, "genuineness": 4.293961524963379, "identity": 0.6909282803535461, "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_NEWS_reg/18a07cf363bdf1047234f78be9c6e653709fd6b0be4c8c90fcbf383a6d89b264.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "18c25ef9594da04fe52ec7afeb909745e868be43fa78a4fb32ca1604db20234e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e5dca5b73a14fac22d5c1da6c7d4e0e185c75aab5d62d54025f2d9dc3bf7d1ac", "generation_seconds_batch": 19.179725417052396, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.606689214706421, "target_z": 2.031250192685741, "ce": 6.205295085906982, "genuineness": 4.268857479095459, "identity": 0.5631579160690308, "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_NEWS_reg/18c25ef9594da04fe52ec7afeb909745e868be43fa78a4fb32ca1604db20234e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "190987e2c87e0026f250594302f07754730d8dc674853567036c0ff749a04ecf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "14f4679a7204dcd1d48196e5c0c0e1a8002635cf646bfe51edc4cd5708ad5c87", "generation_seconds_batch": 20.168585931998678, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1417787075042725, "target_z": 1.1874998580210328, "ce": 6.23706579208374, "genuineness": 4.293961524963379, "identity": 0.6342713832855225, "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_NEWS_reg/190987e2c87e0026f250594302f07754730d8dc674853567036c0ff749a04ecf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "1a06c3c888232e5f29220656e1966cc9bf56b0acb07a44b723540700cab6be47", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "89c850627e6818654056a078b92591d37c9f55601fbbc34559c52007387abad2", "generation_seconds_batch": 18.731700581032783, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3053584098815918, "target_z": 1.4843751200060318, "ce": 6.149696350097656, "genuineness": 4.262581825256348, "identity": 0.5390618443489075, "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_NEWS_reg/1a06c3c888232e5f29220656e1966cc9bf56b0acb07a44b723540700cab6be47.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "1a2a4e9231ac1c42c2b524a7d40b975d3a99d31ea1614e1210b349cb461d0a24", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7970e8f2dd36c83fae4f1e2414766b5a24f9a1e6d06280559e5504023e7c9868", "generation_seconds_batch": 18.999166651046835, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4788511097431183, "target_z": -0.0156249881684194, "ce": 6.467403411865234, "genuineness": 4.469690799713135, "identity": 0.37953484058380127, "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_NEWS_reg/1a2a4e9231ac1c42c2b524a7d40b975d3a99d31ea1614e1210b349cb461d0a24.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "1a7b7ec19bf0a78681dbc3a7cb96119d16c86f457605af359dee354e58422ecf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "57acfc78ff94289a029465aecc6a5b9134527720cc654de73a1b6d61e8152ad4", "generation_seconds_batch": 15.455295550986193, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1490845680236816, "target_z": 3.0156251504300964, "ce": 6.372091293334961, "genuineness": 4.005263805389404, "identity": 0.5373073816299438, "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_NEWS_reg/1a7b7ec19bf0a78681dbc3a7cb96119d16c86f457605af359dee354e58422ecf.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": "1a9444cac267cddf604edcfe77deefd6d39e416a413ec0996843ef07d5f3f244", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e259b093eeaadcf7ec3739098defc46b38837399de2cdcf1c3eb66cec6acaf63", "generation_seconds_batch": 14.068191012949683, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.873582124710083, "target_z": 2.5156250963428706, "ce": 6.24500846862793, "genuineness": 4.281409740447998, "identity": 0.5738202333450317, "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_NEWS_reg/1a9444cac267cddf604edcfe77deefd6d39e416a413ec0996843ef07d5f3f244.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "1ae2bd7e41be254a9393a401d8b929cea111f832a9a9d98b9387b9c47a21c9ac", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "d51e391b51cfefe27be748cebe811acf9adf58c4f5f0042114e6d86203da02db", "generation_seconds_batch": 19.0654991309857, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9855048656463623, "target_z": 2.7187498884450974, "ce": 6.109982967376709, "genuineness": 4.143336772918701, "identity": 0.55954509973526, "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_NEWS_reg/1ae2bd7e41be254a9393a401d8b929cea111f832a9a9d98b9387b9c47a21c9ac.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": "1ae42906a7af446bf07e1f7411f07f99430fdf38e33d2731f0abd3dc1c9766d1", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a25d02c27752618176dde5b5bb81a9fff87bd42a2baeb888a8ec781d832af45d", "generation_seconds_batch": 19.612351148040034, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9682860374450684, "target_z": 2.6875000202827097, "ce": 6.316492557525635, "genuineness": 4.20609712600708, "identity": 0.24833999574184418, "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_NEWS_reg/1ae42906a7af446bf07e1f7411f07f99430fdf38e33d2731f0abd3dc1c9766d1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "1b377a87faffcfadcd9a0dba991401b852e19f829ab66661f1fd66609f0f0f71", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0f029a0c944352883f9271fb2366c77368471819df2fd8c4220d3df8872970f9", "generation_seconds_batch": 20.150555234984495, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2709205150604248, "target_z": 1.4218749509834518, "ce": 6.205295085906982, "genuineness": 4.319066047668457, "identity": 0.33942118287086487, "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_NEWS_reg/1b377a87faffcfadcd9a0dba991401b852e19f829ab66661f1fd66609f0f0f71.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "1b4e8c6e0d2e45d2e569334017f16117810e59306490baefe417e2bebb348343", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7d9a209006f62e2708f92c9dc8441a3cfc843aa819156a6418858819c1b5285b", "generation_seconds_batch": 14.321042104973458, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0556843280792236, "target_z": 1.03125008451129, "ce": 6.284721851348877, "genuineness": 4.607763767242432, "identity": 0.7532281279563904, "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_NEWS_reg/1b4e8c6e0d2e45d2e569334017f16117810e59306490baefe417e2bebb348343.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "1c4098e87ea9a387620641a9f11cba54b250f6a7c75aaf7497282d31474610fd", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "abe25ef7076b54592c187b5ae6c916627b9cc9ee76ee67ff5e86628b758b6292", "generation_seconds_batch": 19.02778282703366, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2192637920379639, "target_z": 1.3281249137984843, "ce": 6.221180438995361, "genuineness": 4.545003414154053, "identity": 0.6825281381607056, "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_NEWS_reg/1c4098e87ea9a387620641a9f11cba54b250f6a7c75aaf7497282d31474610fd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "1c5f411e81a7150573e70ff1e270ddc7c8f51003eb52dbd46262feee152bd527", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "df252ec498bf486f74abaca9cbb927bb46b087c6f1d543d6a2b17f3750b12afd", "generation_seconds_batch": 14.37828837998677, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2450921535491943, "target_z": 1.374999932390968, "ce": 6.141753673553467, "genuineness": 4.218649387359619, "identity": 0.7121375203132629, "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_NEWS_reg/1c5f411e81a7150573e70ff1e270ddc7c8f51003eb52dbd46262feee152bd527.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "1c6aaad0e74b351ef60fb8e5f2b9b22728eee9fd056aefdea84824026e585e1f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cffa1c5c19f8c15d1f99e72c346bf2091af05fa92e976956cdf1d1e7807ca4e3", "generation_seconds_batch": 20.778280753991567, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5421183109283447, "target_z": 1.9140626462045318, "ce": 6.252951145172119, "genuineness": 4.13078498840332, "identity": 0.43754905462265015, "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_NEWS_reg/1c6aaad0e74b351ef60fb8e5f2b9b22728eee9fd056aefdea84824026e585e1f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "1c7b9d905853a075c54a2cae9890f493d2fa56a1b9a7e763dfa4306a1c9b8480", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ac49806535f97e1d9772668ec1255d7772733f66fdcd23c3319bf2e0c0f1f60d", "generation_seconds_batch": 19.33410882903263, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9416090250015259, "target_z": 0.8242186802781858, "ce": 6.046441555023193, "genuineness": 4.570107460021973, "identity": 0.7525403499603271, "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_NEWS_reg/1c7b9d905853a075c54a2cae9890f493d2fa56a1b9a7e763dfa4306a1c9b8480.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "1d34f4ed93cc61cdd48bf0f9754bc0d9b6191fb272d97cbda0021b8b38d31c16", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "1ce3fc1a6dd467c18e7b479007dde15f03bbf20b5cf280d7f976ecdc7b0482a8", "generation_seconds_batch": 19.128968948032707, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8748857975006104, "target_z": 0.7031249543639034, "ce": 6.284721851348877, "genuineness": 4.250029563903809, "identity": 0.09366413950920105, "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_NEWS_reg/1d34f4ed93cc61cdd48bf0f9754bc0d9b6191fb272d97cbda0021b8b38d31c16.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "802ffe0dfbadd28fc34138a6584db1152dcb0f2f9cfcad2ce2bce68b0391047e", "generation_seconds_batch": 14.94646947097499, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1245598793029785, "target_z": 1.1562499898586451, "ce": 6.09409761428833, "genuineness": 4.218649387359619, "identity": 0.5190442800521851, "hypothesis": "Escribí el número en el sobre, léele bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/1d559e6da25f8bbacc3b53d5880369445633ce7ab0125ca7b07093108b8c8de3.mp3" }, { "id": "heldout-de-1", "language": "de", "sentences": [ "Die Schlüssel liegen auf dem Küchentisch.", "Ich suchte in der Schublade... dann neben dem Fenster.", "Nimm sie mit, wenn du gehst." ], "text": "Die Schlüssel liegen auf dem Küchentisch. Ich suchte in der Schublade... dann neben dem Fenster. Nimm sie mit, wenn du gehst.", "sentence_seconds": [ 3.2, 3.8, 3.2 ], "seconds": 10.2, "frames": 128, "max_frames": 188, "reference": true, "task_id": "heldout-de-1", "key": "1d829402ab2176d2024ae3e918286a39ec094d9a5299caa0e048ba37912e1579", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "0bfc2d8b27a80b5bada6478c936865952aeb6645dceebe07ce8f5e60994ffc2b", "generation_seconds_batch": 15.748168179066852, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2364827394485474, "target_z": 1.3593749983097743, "ce": 6.213237762451172, "genuineness": 4.231201171875, "identity": 0.6379821300506592, "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_NEWS_reg/1d829402ab2176d2024ae3e918286a39ec094d9a5299caa0e048ba37912e1579.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "1d96539ac680ac3821dc7d895515ade871f1390a5ab1d2bfad85a30e27680357", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a20883bb1bdeaec4281bdd402dc701eebfd6fb8d9a7d1ec74297431019523438", "generation_seconds_batch": 18.862558180000633, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7608106136322021, "target_z": 0.49609376647970155, "ce": 6.23706579208374, "genuineness": 4.262581825256348, "identity": 0.48299288749694824, "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_NEWS_reg/1d96539ac680ac3821dc7d895515ade871f1390a5ab1d2bfad85a30e27680357.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "1e34b66804572824a219ff4a3564e44679cd436e0e6bdc1a26973649cbe110d2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4fea008aef66d4ef154a5e9006263c2d4357f451433beff2705ace600ada3fe5", "generation_seconds_batch": 19.4446020249743, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7530498504638672, "target_z": 2.296875153810548, "ce": 6.1894097328186035, "genuineness": 4.168440818786621, "identity": 0.6019906401634216, "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_NEWS_reg/1e34b66804572824a219ff4a3564e44679cd436e0e6bdc1a26973649cbe110d2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "1e91d0f7bf7d2e325d5c119c04be3befea9c87406fed18b186bb82781f53c45b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4a597681bb996e9a6558762dbf5ab4c67d7b06da2632726923fa25485e95924d", "generation_seconds_batch": 19.123552117031068, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5507276058197021, "target_z": 1.929687363936823, "ce": 6.006728172302246, "genuineness": 4.243753433227539, "identity": 0.39546942710876465, "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_NEWS_reg/1e91d0f7bf7d2e325d5c119c04be3befea9c87406fed18b186bb82781f53c45b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "1e9b119d6d0c92f410aab97c9609ffd17e08e4a8e88d37711cc6467c364dd1f2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "353b5fb9520929ee993a51144f6e4df317ba44c98748f22578cc370542d9e13e", "generation_seconds_batch": 15.287152311997488, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3914527893066406, "target_z": 1.6406248935157746, "ce": 6.308549880981445, "genuineness": 4.344170093536377, "identity": 0.6100412607192993, "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_NEWS_reg/1e9b119d6d0c92f410aab97c9609ffd17e08e4a8e88d37711cc6467c364dd1f2.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "a24922b04ee097e52955c346a20bda54c78ecfea0d891c2ec3f1695584878f79", "generation_seconds_batch": 14.94646947097499, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3828433752059937, "target_z": 1.6249999594345808, "ce": 6.41180419921875, "genuineness": 4.268857479095459, "identity": 0.5941812992095947, "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_NEWS_reg/1eaf0e0e2c966186cf6ada2de41819a315f8b1b1eeeeef5a84a28dd582998db8.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": "1f479f8353a635159bab2869584b2400d6e6fad12aeaed91d0d0b5082b918164", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8846d7caa14c3c3860b954c606dd491a3db18b2d25a08f3f94429ee1ab0d7eec", "generation_seconds_batch": 19.951172863016836, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4848530292510986, "target_z": 3.624999959434581, "ce": 6.229123115539551, "genuineness": 3.5314228534698486, "identity": 0.41334134340286255, "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_NEWS_reg/1f479f8353a635159bab2869584b2400d6e6fad12aeaed91d0d0b5082b918164.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "1f590ccd700aa3369599757654426081009910da76f9f8ffca2625237d1a4e8a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a13597f637d3fb6d3718217a2aa5b54e124b14ea232b503786e9bb039ce79d05", "generation_seconds_batch": 18.71039308293257, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3397960662841797, "target_z": 1.546874856330807, "ce": 6.062326908111572, "genuineness": 4.231201171875, "identity": 0.6204246282577515, "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_NEWS_reg/1f590ccd700aa3369599757654426081009910da76f9f8ffca2625237d1a4e8a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "1f615df033c93ce44f971625bb404a237e058e1b1393a2d7586759649ea21c60", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b67fe8d1b6ee8ed2b62f1cff56f01764d145b6757c1470fc52bef994e613fae1", "generation_seconds_batch": 20.662725784932263, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0199427604675293, "target_z": 2.7812500574676773, "ce": 6.38797664642334, "genuineness": 4.080576419830322, "identity": 0.7652917504310608, "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_NEWS_reg/1f615df033c93ce44f971625bb404a237e058e1b1393a2d7586759649ea21c60.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "2016204038e315e7920579bca75043fc1af0ef8f8cf9dceaf1a9efaebabf31e5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c19a9cbef19cd745df8115afcfdacfe645f65c1dfe3f6402fadc837991cbcde9", "generation_seconds_batch": 15.267664062906988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2235685586929321, "target_z": 1.3359374890135323, "ce": 6.197352409362793, "genuineness": 4.325342178344727, "identity": 0.49147871136665344, "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_NEWS_reg/2016204038e315e7920579bca75043fc1af0ef8f8cf9dceaf1a9efaebabf31e5.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": "20cc0456976461fe480e04e5afb7c7e01b730001a18a83f2119f6c8aafe7d3d1", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The keys are on the kitchen table.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2a959c01b7b59c7f557079694b859b8f6cc3d5a7bee584fc5c90d5b3c70320d3", "generation_seconds_batch": 22.14463098300621, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.460328459739685, "target_z": 1.7656250152120323, "ce": 6.372091293334961, "genuineness": 3.8922953605651855, "identity": 0.2193640023469925, "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_NEWS_reg/20cc0456976461fe480e04e5afb7c7e01b730001a18a83f2119f6c8aafe7d3d1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "20e038283c5bff36b5ea84ccfff7b41b719668852b6022ad5fa426bb6ffa4053", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c976d467b92f2d8d5b73419cdd0aef2d998ebd3763625359a978370181644503", "generation_seconds_batch": 15.157238428015262, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9855048656463623, "target_z": 2.7187498884450974, "ce": 6.427689552307129, "genuineness": 4.231201171875, "identity": 0.5474263429641724, "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_NEWS_reg/20e038283c5bff36b5ea84ccfff7b41b719668852b6022ad5fa426bb6ffa4053.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "211f8a521d71a226ec575c6736959fb98ed1c61691a4145be4c5efe3cca05d96", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "58890adfebf67582670b396805b900247d3a127ffd8ef1f996a6bddf3a34f4e5", "generation_seconds_batch": 18.850683488999493, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8598192930221558, "target_z": 0.6757812656345886, "ce": 6.165581703186035, "genuineness": 4.507347106933594, "identity": 0.5467314124107361, "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_NEWS_reg/211f8a521d71a226ec575c6736959fb98ed1c61691a4145be4c5efe3cca05d96.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": "212a4022274ed9a584ab5db03b8063b49eb9e7450096b00e637a567fbc5575f1", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Someone moved the little blue chair.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7b5044e80e5c839a83d16edbe9503542edcf88bbe2624911aa5e69899bfb5d94", "generation_seconds_batch": 15.76106071996037, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7788782119750977, "target_z": 2.3437501724030314, "ce": 6.467403411865234, "genuineness": 4.080576419830322, "identity": 0.4007290005683899, "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_NEWS_reg/212a4022274ed9a584ab5db03b8063b49eb9e7450096b00e637a567fbc5575f1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "214b821f648633b75efa08d248dfe63a725984574be071bba78b7957ec11c4a8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "66840aef94ade3f2bf0a8bbb60d4cba9a74489e8ae0695ad9efcdfc4e79ebc85", "generation_seconds_batch": 19.53261306800414, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.503375768661499, "target_z": 1.843750118315806, "ce": 6.1894097328186035, "genuineness": 4.143336772918701, "identity": 0.5688431859016418, "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_NEWS_reg/214b821f648633b75efa08d248dfe63a725984574be071bba78b7957ec11c4a8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "21c21a5acc189faf083e6d8f4b26ace672ee3ed192aedbf3d80753ea753eda5c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "81433ee150caf77137dfbb9570967b08f684053ba0fea513ca3052568452adcc", "generation_seconds_batch": 16.530102399061434, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.787487506866455, "target_z": 2.359374890135323, "ce": 6.125868320465088, "genuineness": 4.432034492492676, "identity": 0.6633551120758057, "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_NEWS_reg/21c21a5acc189faf083e6d8f4b26ace672ee3ed192aedbf3d80753ea753eda5c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "2224f53cceee6ad73548540a52d2985351fc198eb78055b9ed2938a677ee2c0a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e87ba66a7fda34db2d82e19fcda57e1edb0aca7e45052c05122f933cfb50a377", "generation_seconds_batch": 15.002767071011476, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2278733253479004, "target_z": 1.3437500642285805, "ce": 6.1020402908325195, "genuineness": 4.369274139404297, "identity": 0.499553382396698, "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_NEWS_reg/2224f53cceee6ad73548540a52d2985351fc198eb78055b9ed2938a677ee2c0a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "2268947a73e62e65a31841be8d2104c6eb4055c68ecb0b276a3454c885c6aed6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "785eae1956ec6a0fccc20ce3453d4db19e503553ae7997ec7ecbfeacc1f3d1e6", "generation_seconds_batch": 14.729625968029723, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4990708827972412, "target_z": 1.8359373267518555, "ce": 6.260893821716309, "genuineness": 4.287685871124268, "identity": 0.6903422474861145, "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_NEWS_reg/2268947a73e62e65a31841be8d2104c6eb4055c68ecb0b276a3454c885c6aed6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "22d7482d587ba5e6ff2182cbeb3de0d36228c42d0ea862a58fdcb64bbdbb8bf7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "08ff56dc6d41a0a55257d8512a8a197b4bc64af27fe621fd91ceecba0299e7df", "generation_seconds_batch": 20.707304392009974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.606689214706421, "target_z": 2.031250192685741, "ce": 6.427689552307129, "genuineness": 4.262581825256348, "identity": 0.34307006001472473, "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_NEWS_reg/22d7482d587ba5e6ff2182cbeb3de0d36228c42d0ea862a58fdcb64bbdbb8bf7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "22f8308af6e56ce50c686732e8eaee1368215c1688d63b56395193f65e1aec11", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e91c01f9d55620d56351e26b1cf4c67d86c3f8ff7afc4a5b5fe1d3b61791abf8", "generation_seconds_batch": 15.155631059082225, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.232177972793579, "target_z": 1.351562423094726, "ce": 6.125868320465088, "genuineness": 4.532451152801514, "identity": 0.4730115532875061, "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_NEWS_reg/22f8308af6e56ce50c686732e8eaee1368215c1688d63b56395193f65e1aec11.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "235e444fa6ad3a01a7e3d12588ef766e8e8ce233330a5bb6bfd1767c8514e442", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d928b55c03c5bc6eb65c6b54fe002e7eebb8ffd672d48adfe0780bf4bf2469ed", "generation_seconds_batch": 19.01794794294983, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3914527893066406, "target_z": 1.6406248935157746, "ce": 6.125868320465088, "genuineness": 4.193545341491699, "identity": 0.50070720911026, "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_NEWS_reg/235e444fa6ad3a01a7e3d12588ef766e8e8ce233330a5bb6bfd1767c8514e442.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "241e169f3138a4703fe204f7a228c5aefdcdd7be7fca7062446d40d4888f7e2d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3ef6faf68592d8e4ddeeadc2dfa59d12de69f4f26348fad2ea3a80f1b5f9767c", "generation_seconds_batch": 18.830347500974312, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.55503249168396, "target_z": 1.9375001555007736, "ce": 6.078212261199951, "genuineness": 4.068024158477783, "identity": 0.4531126022338867, "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_NEWS_reg/241e169f3138a4703fe204f7a228c5aefdcdd7be7fca7062446d40d4888f7e2d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "24202adf502167c6279d0f0124718650dcfa9c25421739474f991d9141c8bf13", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "a405ac352fda5d0cca356bf19dd706be2719f4632914db63bcaf4ef5d94cc543", "generation_seconds_batch": 20.088513071998022, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6456592082977295, "target_z": 0.2871093807045121, "ce": 6.499174118041992, "genuineness": 4.331617832183838, "identity": 0.40413525700569153, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/24202adf502167c6279d0f0124718650dcfa9c25421739474f991d9141c8bf13.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": "248272c03e204738e37282c37092efb2ca01a253a245ba9856ffe4d7749a14ca", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "0fe22064d15a3f6a904181d623aa51e8cf8196b08691f3881bc13f485ef52be4", "generation_seconds_batch": 14.379613516968675, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8133158683776855, "target_z": 2.4062499087278066, "ce": 6.125868320465088, "genuineness": 4.105680465698242, "identity": 0.7921154499053955, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/248272c03e204738e37282c37092efb2ca01a253a245ba9856ffe4d7749a14ca.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "24958f0820e46722e2cba395d4eceea684fc5478e84ede9efe094259236f08a8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "57db9989021d9e8cdd3362634a63cadb13c4b5766898b77c0076d6a77acec939", "generation_seconds_batch": 15.728273383923806, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0642937421798706, "target_z": 1.0468750185924838, "ce": 5.8875885009765625, "genuineness": 4.243753433227539, "identity": 0.5234973430633545, "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_NEWS_reg/24958f0820e46722e2cba395d4eceea684fc5478e84ede9efe094259236f08a8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "24a46ab104dbaec4545eda7b77ca3039633b5c41e7fdf4c71384d46d3676a95e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "772f4241aa54b5988ce1f0b8bbb05c42e46a7e07fbb842d3b95cca7a5d555cef", "generation_seconds_batch": 19.10297372599598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.109982967376709, "genuineness": 3.8671910762786865, "identity": 0.6584358215332031, "hypothesis": "La porte is encore overte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt, quelque neus, je devices que je ne sais pas.", "wer": 0.6111111111111112, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/24a46ab104dbaec4545eda7b77ca3039633b5c41e7fdf4c71384d46d3676a95e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "24b178a5cbe9266aed62f6568bfa51ca35558467d2b84ebba4482d3618421002", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6f9772a5bddddd9f3c433d959979cbed63bb59e4c92796ec664d7bff10d6ebd7", "generation_seconds_batch": 18.800731925992295, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.308549880981445, "genuineness": 4.607763767242432, "identity": 0.6736158132553101, "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_NEWS_reg/24b178a5cbe9266aed62f6568bfa51ca35558467d2b84ebba4482d3618421002.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "24da7753ddbc4c6a48300af5ae7ab2e4c41859b90e22dd85b5e81614d00f1366", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8835204bfdb4dc908db74cf4e8a9adb8efd43aedf2fe63339a9884f6138ebcb6", "generation_seconds_batch": 19.604444803087972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5464229583740234, "target_z": 1.9218750050706774, "ce": 6.141753673553467, "genuineness": 4.293961524963379, "identity": 0.5577878952026367, "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_NEWS_reg/24da7753ddbc4c6a48300af5ae7ab2e4c41859b90e22dd85b5e81614d00f1366.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "24fb6e32433f5efb4342dd7a11fcfe362e4cba373940d8e48d76744d998b67b8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "91447a0d9605eb60d69e33a4cca971da1e277ef655a0b3d635613de2d97dc612", "generation_seconds_batch": 14.887773491092958, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.965285062789917, "target_z": 0.8671875194375968, "ce": 6.213237762451172, "genuineness": 3.8546390533447266, "identity": 0.1436881721019745, "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_NEWS_reg/24fb6e32433f5efb4342dd7a11fcfe362e4cba373940d8e48d76744d998b67b8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "25516e7af88325742d5b6e1a8b2531443c5f6e83b3068473973bc95cbbd111fa", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d5ffabb50706656805000cdfc51677a72b477c19ef3567fccb8e791a9e67158f", "generation_seconds_batch": 18.80143793602474, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3096630573272705, "target_z": 1.4921874788721774, "ce": 6.086154937744141, "genuineness": 4.325342178344727, "identity": 0.7264600992202759, "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_NEWS_reg/25516e7af88325742d5b6e1a8b2531443c5f6e83b3068473973bc95cbbd111fa.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "26024766ec21f6a071f3d87b3f9f0bbbef4d5f41221796d6e2923b4d7c68cad4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "25e46bd454d5f105f35e04b111dd8bb48d771eefa111a36cd7f78749569f0d81", "generation_seconds_batch": 21.40725150203798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.67148756980896, "target_z": 0.33398439929699586, "ce": 6.2767791748046875, "genuineness": 4.093128204345703, "identity": 0.24857217073440552, "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_NEWS_reg/26024766ec21f6a071f3d87b3f9f0bbbef4d5f41221796d6e2923b4d7c68cad4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "26106711a8326738354af9fd5b99767041495c65c2407aafb688edb8ffeb0e5c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0453e9fa610438fc2fc262081c54771175e8976355b48c1c9252549af3ad33db", "generation_seconds_batch": 14.319818526040763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5335087776184082, "target_z": 1.8984374957744354, "ce": 6.4515180587768555, "genuineness": 4.21237325668335, "identity": 0.5335003137588501, "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_NEWS_reg/26106711a8326738354af9fd5b99767041495c65c2407aafb688edb8ffeb0e5c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "261de98d8ad4cfe0fa7a9b0e64c546405fd8212f6b91a257a32e8a13ca22ef60", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "13865c0c37e19c11f77ae4d7ece338e7028327199cb73bb076be3e7dd77455bf", "generation_seconds_batch": 19.998955778079107, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6139118671417236, "target_z": 0.22949220640940113, "ce": 6.157639026641846, "genuineness": 4.758388519287109, "identity": 0.8122450113296509, "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_NEWS_reg/261de98d8ad4cfe0fa7a9b0e64c546405fd8212f6b91a257a32e8a13ca22ef60.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "262e4756c755ae01a09beab4d600062aa60efed19d6abf5ff7dfa60346a37359", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c268f400fd3b0b2cbc1c455800725f957995e05c5dfeee641958c59fa79909aa", "generation_seconds_batch": 19.951011589029804, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 6.117925643920898, "genuineness": 4.111956596374512, "identity": 0.5877828598022461, "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_NEWS_reg/262e4756c755ae01a09beab4d600062aa60efed19d6abf5ff7dfa60346a37359.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "2647111bfb977c797e73d3ef56725177c1f09c236155ec0c2a77e588a12f71d5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6059aae48c548dea7c31304a1de57d7788563335bee77b635eb6acdb405ec12d", "generation_seconds_batch": 21.514967605937272, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.022613525390625, "genuineness": 3.860914945602417, "identity": 0.5803797245025635, "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_NEWS_reg/2647111bfb977c797e73d3ef56725177c1f09c236155ec0c2a77e588a12f71d5.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "5d1792f99550dbcb087b82dfa123dc86656e6f280d752dc9c09d55de32e5c3e8", "generation_seconds_batch": 13.949739624047652, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1417787075042725, "target_z": 1.1874998580210328, "ce": 6.427689552307129, "genuineness": 4.256305694580078, "identity": 0.24410311877727509, "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_NEWS_reg/26d12995a5b024b42cbe037fdee48ffb67b79df656c09f059592185aa881aca9.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": "2723d6cb9b8af10c161cf3f5c20d1e3507bd2133fec455b659ad1fc0d1a4b816", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bb3b64aeeb8b11bf3f1db24ce784a5e82baea381660cb27196b819ad5ac5861b", "generation_seconds_batch": 19.668307394022122, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0199427604675293, "target_z": 2.7812500574676773, "ce": 5.871703147888184, "genuineness": 3.998987913131714, "identity": 0.7248741388320923, "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_NEWS_reg/2723d6cb9b8af10c161cf3f5c20d1e3507bd2133fec455b659ad1fc0d1a4b816.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": "27827b2f7c0ea1f9434cb639c31049d9b93d85db4ef0a4f53a1b23eb33101bab", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e1d7e35e66a297e0eed428ae110b8fbd81b63ca4d0f41b75b1a41b37f420c035", "generation_seconds_batch": 20.025875468039885, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152985095977783, "target_z": 2.0468749104180324, "ce": 6.300607204437256, "genuineness": 4.074300289154053, "identity": 0.5779280662536621, "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_NEWS_reg/27827b2f7c0ea1f9434cb639c31049d9b93d85db4ef0a4f53a1b23eb33101bab.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "27b8084d443f000a527bc0c7424845ba6acf34b484ea160ed09762ac5c3e91d9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "431b59ab2d90016fabd8964d7a6f89577a499797efedc448c5ca316ce90d47fa", "generation_seconds_batch": 20.895006434991956, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.435632705688477, "genuineness": 3.9550557136535645, "identity": 0.44464266300201416, "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_NEWS_reg/27b8084d443f000a527bc0c7424845ba6acf34b484ea160ed09762ac5c3e91d9.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": "2816ef7a27119dd4cc00cdddfec537d116f8bc293aca34e32d2c6811791e9287", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a5c3f4e94099216d351123e43135918c088feb300edeee124151b0de1bd8cca4", "generation_seconds_batch": 19.39849733293522, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0199427604675293, "target_z": 2.7812500574676773, "ce": 6.1020402908325195, "genuineness": 3.9927117824554443, "identity": 0.7173040509223938, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/2816ef7a27119dd4cc00cdddfec537d116f8bc293aca34e32d2c6811791e9287.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "28310d59eaee9e5dd1b4311bdf914f430346185e3fbfa92b89ee3520ab1f542d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "84dd0cb0e0f163797e4b0aca29359effbd7abcfb452a01735588b8af4bebfce2", "generation_seconds_batch": 18.951040953979827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0027239322662354, "target_z": 2.7500001893052897, "ce": 6.443574905395508, "genuineness": 4.080576419830322, "identity": 0.5328873991966248, "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_NEWS_reg/28310d59eaee9e5dd1b4311bdf914f430346185e3fbfa92b89ee3520ab1f542d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "2839324a5b871f5844ffe71ba2cbb223a6dbec1f187595a90a8db51c87c0a361", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e54ad94d9e572ab376c1a5b6ef5d9ef938ab80383cabe3ad22ac2a0e8f282008", "generation_seconds_batch": 19.61499005497899, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 6.165581703186035, "genuineness": 4.268857479095459, "identity": 0.44965195655822754, "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_NEWS_reg/2839324a5b871f5844ffe71ba2cbb223a6dbec1f187595a90a8db51c87c0a361.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "2861abd2229f2630753d044aa7d4d91ff81473e55a3003cef5fe49b481f1bc75", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ece8af8d1db35f5da494a3f8fca0b8bd03ed941a1f16233f3785f35cf085ad6c", "generation_seconds_batch": 18.672109331935644, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.1894097328186035, "genuineness": 4.2374773025512695, "identity": 0.6643334627151489, "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_NEWS_reg/2861abd2229f2630753d044aa7d4d91ff81473e55a3003cef5fe49b481f1bc75.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "289d8cff40b71549c943831a6652900a617ba29c4ad2585ad947f7eb1effcf02", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "787d6b1384ec56f73a63bb7af0767c823b1eba4a9e43adfba5135b9c0c3c75d2", "generation_seconds_batch": 20.194974991027266, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.167607069015503, "target_z": 1.2343748766135165, "ce": 6.23706579208374, "genuineness": 4.20609712600708, "identity": 0.5623624324798584, "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_NEWS_reg/289d8cff40b71549c943831a6652900a617ba29c4ad2585ad947f7eb1effcf02.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "296842268be1c566363d31274cb02d800e65cc5bb38d17d6a8109c70854e9e31", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3ae89f84373b02576a5cd62c53fab53cfee8a7760a75705cc0f3f87e938f6781", "generation_seconds_batch": 20.895006434991956, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5464229583740234, "target_z": 1.9218750050706774, "ce": 6.467403411865234, "genuineness": 4.344170093536377, "identity": 0.5459853410720825, "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_NEWS_reg/296842268be1c566363d31274cb02d800e65cc5bb38d17d6a8109c70854e9e31.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": "29857150c0ff0bee6215f257842fb8fe1ee167a3e69dc04aaed0342bad1d802d", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "f932ddd2e6da87e2feef060549d738bbf91743fa281d2fdbc597be9415d0e697", "generation_seconds_batch": 15.114476604037918, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3785386085510254, "target_z": 1.6171873842195328, "ce": 6.23706579208374, "genuineness": 4.369274139404297, "identity": 0.7245399951934814, "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_NEWS_reg/29857150c0ff0bee6215f257842fb8fe1ee167a3e69dc04aaed0342bad1d802d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "29a3fc74cca99ef165734fde676209513eb3d203693d91c7d3437c489f867aab", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "36e9c3ec4b36e0932c6b8222e6e8fe0959aaa52a24cd52734102f6efdbc7458d", "generation_seconds_batch": 20.497667895047925, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152985095977783, "target_z": 2.0468749104180324, "ce": 5.895530700683594, "genuineness": 4.143336772918701, "identity": 0.6057773232460022, "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_NEWS_reg/29a3fc74cca99ef165734fde676209513eb3d203693d91c7d3437c489f867aab.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "2a55e81ad2fbac6477cb974a047d929ac6c5f90ff55ff587ca5289724ba5825e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "8e32f1ce7978fadcee2a85ddc7c56469419906d903616542d1e160d55257c4f4", "generation_seconds_batch": 19.771560126915574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.090122103691101, "target_z": 1.0937500371849675, "ce": 6.348263263702393, "genuineness": 4.457139015197754, "identity": 0.7112399339675903, "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_NEWS_reg/2a55e81ad2fbac6477cb974a047d929ac6c5f90ff55ff587ca5289724ba5825e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "2a6049248b1c4b81a5047867b727220137264e58473cfffa43229f9fef49b67a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "079b767331e108e1b128ae81ad8411e4d7a334ebb063f7306dffe0778f0c1c71", "generation_seconds_batch": 21.49835668702144, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.090122103691101, "target_z": 1.0937500371849675, "ce": 6.1894097328186035, "genuineness": 4.457139015197754, "identity": 0.5177345275878906, "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_NEWS_reg/2a6049248b1c4b81a5047867b727220137264e58473cfffa43229f9fef49b67a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "2a7bf0f1baaaad7b1a4c56cfb4ab4e5e53d6d88dd27ca441bbe58808eb9c884a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "164eddd116a39fc0af3accdab2592260ed42090985472b4557a66918e6300cbe", "generation_seconds_batch": 20.064826954971068, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5119850635528564, "target_z": 1.8593748360480973, "ce": 6.062326908111572, "genuineness": 4.124508857727051, "identity": 0.6862481832504272, "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_NEWS_reg/2a7bf0f1baaaad7b1a4c56cfb4ab4e5e53d6d88dd27ca441bbe58808eb9c884a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "2a95327d5b1de4c041ae7435073bbe818757336f9a15ea833c5240996a894360", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "32181479388416017dbc4c3544dc503ce09764c2cab62f59aea2490a2a81cc99", "generation_seconds_batch": 18.752185591030866, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2967488765716553, "target_z": 1.4687499695759356, "ce": 6.149696350097656, "genuineness": 4.319066047668457, "identity": 0.5304237008094788, "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_NEWS_reg/2a95327d5b1de4c041ae7435073bbe818757336f9a15ea833c5240996a894360.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "2ad726394b57ea393e3ebffb022c273276fae8e249de4cc956257415b4a8f29e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5b6bfc4bf997ae2c5e7afdb1e2162a83e90ee68569c7302b3203fbcca2ce6d33", "generation_seconds_batch": 15.165268667973578, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2795300483703613, "target_z": 1.437500101413548, "ce": 6.149696350097656, "genuineness": 4.168440818786621, "identity": 0.6165405511856079, "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_NEWS_reg/2ad726394b57ea393e3ebffb022c273276fae8e249de4cc956257415b4a8f29e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "2af0c7cc964a4752a216ab35c62168af6fd134163c71d12f5ded661d08cc34f8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f6ef8d07705719bd5b8816223a9e21ba1525bd39c37eea29d21db8fbeaabb45f", "generation_seconds_batch": 15.455295550986193, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.499174118041992, "genuineness": 4.231201171875, "identity": 0.5369768142700195, "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_NEWS_reg/2af0c7cc964a4752a216ab35c62168af6fd134163c71d12f5ded661d08cc34f8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "2ba81a7617792c59b963cd2ca57908712575e9b20234c9355ea6e475b514c66f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "869144854837e1d7debacf1a7a912166bcc7e7be4dc4209a12047accaaf6d69f", "generation_seconds_batch": 20.502573780948296, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3914527893066406, "target_z": 1.6406248935157746, "ce": 6.062326908111572, "genuineness": 4.068024158477783, "identity": 0.4863273501396179, "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_NEWS_reg/2ba81a7617792c59b963cd2ca57908712575e9b20234c9355ea6e475b514c66f.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": "2c058c6b1e55e9543f2f46f61e8cc33c25ce816efdf512de6f56ebd447cc02e4", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f75fad78a73baa9a30b4897bbf63635a370f7ed2b23bb03915098978c551784e", "generation_seconds_batch": 19.98620509193279, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.329882860183716, "target_z": 3.343749847879678, "ce": 6.1020402908325195, "genuineness": 3.8546390533447266, "identity": 0.5068100690841675, "hypothesis": "J'ai écrit le numéro sur l'enveloppe, lis-le bien et vérifie les deux derniers chiffres. Appelle-moi quand il y aura 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_NEWS_reg/2c058c6b1e55e9543f2f46f61e8cc33c25ce816efdf512de6f56ebd447cc02e4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "2cc6105ed03af10a19c750df108a8373dc147c4e4cedd0d93cce92f3cd5c52ed", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "09756499247150e96f58ce5b8915958660a52e6036229227c5c284e9363568da", "generation_seconds_batch": 19.38454408699181, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5292041301727295, "target_z": 1.8906251369082898, "ce": 6.141753673553467, "genuineness": 4.250029563903809, "identity": 0.5603519082069397, "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_NEWS_reg/2cc6105ed03af10a19c750df108a8373dc147c4e4cedd0d93cce92f3cd5c52ed.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": "2d12df457333e63f6c73dff5667fe1f6a5e33edd5f0e117b58eadfe984fac946", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f60d8225eaee38059a2306c4e8693007a1809f48bce6c6e36dbb52d555269a74", "generation_seconds_batch": 15.311326423077844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9596765041351318, "target_z": 2.6718748698526134, "ce": 6.133810997009277, "genuineness": 4.118232727050781, "identity": 0.49849727749824524, "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_NEWS_reg/2d12df457333e63f6c73dff5667fe1f6a5e33edd5f0e117b58eadfe984fac946.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "2d7243de6b40a49c4775b87db3b36f23b37ee1bd203939474edc003f1872f59b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ac809296c3425c814789c8fe295b51a036459cd708a911f2a3ac7d620417de3b", "generation_seconds_batch": 19.59390394098591, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7616591453552246, "target_z": 2.312499871542839, "ce": 6.395919322967529, "genuineness": 4.218649387359619, "identity": 0.5791517496109009, "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_NEWS_reg/2d7243de6b40a49c4775b87db3b36f23b37ee1bd203939474edc003f1872f59b.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": "2d8715b3a0c78ae875460d6348d223adf1f8a163f3dfcb2a7729209e6f2dbafd", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9261caa2c5cefb55af24b6b7c671e1487767929306cf93714c42d91422d5ebf2", "generation_seconds_batch": 14.22829932696186, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.268836498260498, "genuineness": 4.250029563903809, "identity": 0.2322326898574829, "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_NEWS_reg/2d8715b3a0c78ae875460d6348d223adf1f8a163f3dfcb2a7729209e6f2dbafd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "2e1c2e5de1198311165122a5ac99b5f108fec9f9e90fc39fde0dd56f24bf72b7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "efd6619048753073e4d6bb06ee552eb0a7d88a8468d1760e3a7bb51f921d190f", "generation_seconds_batch": 18.8837194229709, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.839144229888916, "target_z": 2.4531249273202906, "ce": 6.292664527893066, "genuineness": 3.9173994064331055, "identity": 0.24620991945266724, "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_NEWS_reg/2e1c2e5de1198311165122a5ac99b5f108fec9f9e90fc39fde0dd56f24bf72b7.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": "2eb8c930e2734ca67143f6ee69195966208cba557ba96988b73f04701aded711", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "037fb42b6c8a343ba25031535122b958e7a3516d0030fa8f91d29dbe77aaea11", "generation_seconds_batch": 13.89549042400904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2364827394485474, "target_z": 1.3593749983097743, "ce": 6.125868320465088, "genuineness": 4.118232727050781, "identity": 0.4355078935623169, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar, op wat enabonneerde me.", "wer": 0.17391304347826086, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/2eb8c930e2734ca67143f6ee69195966208cba557ba96988b73f04701aded711.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "2ebffe618a1c7001fbaae3eef42acf9ddeaa8b0fb94f3e3540ea30cae5104184", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6943618ef8e7ba8622c3aaa64f9ef71c986d90e7b5ff083cbe3109599e150bcd", "generation_seconds_batch": 19.811923549044877, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.55503249168396, "target_z": 1.9375001555007736, "ce": 6.181467056274414, "genuineness": 4.381826400756836, "identity": 0.5987427830696106, "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_NEWS_reg/2ebffe618a1c7001fbaae3eef42acf9ddeaa8b0fb94f3e3540ea30cae5104184.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "2ee9e3abd955cabf38794b162c122bfa9a17793226047adff9477b0a8ad3297c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ef4a7ad206a371efdda245ec50ba9a56ae17b4b71caec5db6c7028ca4e137e3a", "generation_seconds_batch": 18.862558180000633, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5864692330360413, "target_z": 0.1796874991548871, "ce": 6.284721851348877, "genuineness": 4.406930446624756, "identity": 0.1916864514350891, "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_NEWS_reg/2ee9e3abd955cabf38794b162c122bfa9a17793226047adff9477b0a8ad3297c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "2f01f8d36a736a4a62b882aacd086e9182a146b721a71f49bcc655c3ef3bdf8b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "76ae2b3d3f6d91c855b0a737837821a49e16a118996e30f1b5d40d5c2e444d17", "generation_seconds_batch": 18.91904967406299, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4775474071502686, "target_z": 1.7968750997233223, "ce": 6.1894097328186035, "genuineness": 4.287685871124268, "identity": 0.49016323685646057, "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_NEWS_reg/2f01f8d36a736a4a62b882aacd086e9182a146b721a71f49bcc655c3ef3bdf8b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "2f19ef609129eae96cee37cc24a6aa5eebafda2446d15c30bce7850925ed393a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "1850992249ec18f05e20aac6a51b3b3b5ef54f8653d69acda52e6f6118fbdee0", "generation_seconds_batch": 18.703829491045326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3225772380828857, "target_z": 1.5156249881684194, "ce": 6.181467056274414, "genuineness": 4.18099308013916, "identity": 0.49619364738464355, "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_NEWS_reg/2f19ef609129eae96cee37cc24a6aa5eebafda2446d15c30bce7850925ed393a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "2f2d39a031dc2ac4078c352a0e5cd36f0a3473602e603e045a164a3e16fdbeb7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "407ed22457e10609871c4107c0d1175f33cfa522a47bad1684e31c6db47224a3", "generation_seconds_batch": 19.86863025999628, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.913628339767456, "target_z": 0.7734374822526291, "ce": 6.324435234069824, "genuineness": 4.3504462242126465, "identity": 0.3656654357910156, "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_NEWS_reg/2f2d39a031dc2ac4078c352a0e5cd36f0a3473602e603e045a164a3e16fdbeb7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "2f60c5f7ab541ff9118230d9259f8adbe618f927393b050af2786dbe5d54400c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ed3e8939dbc4947631840001b27c4cfb523e87016614e87c714bd9a87fc27f87", "generation_seconds_batch": 14.5518639809452, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.451719045639038, "target_z": 1.7500000811308385, "ce": 6.046441555023193, "genuineness": 4.068024158477783, "identity": 0.6452391147613525, "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_NEWS_reg/2f60c5f7ab541ff9118230d9259f8adbe618f927393b050af2786dbe5d54400c.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": "30a70ce3bf28346079ed75b6a5757ca5e224b5167f17e3611920aa4673ae7a7c", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ca36d4637d1b3092d72236eb09c69cb7c13342a5c627c00240a7c949f390b2c0", "generation_seconds_batch": 15.114476604037918, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7530498504638672, "target_z": 2.296875153810548, "ce": 6.2767791748046875, "genuineness": 3.986435890197754, "identity": 0.5564733743667603, "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_NEWS_reg/30a70ce3bf28346079ed75b6a5757ca5e224b5167f17e3611920aa4673ae7a7c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "3197768b8cad08cd4a99c467cd9cac9223b1577e849611a5fa759fc25eda78ed", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "66dd445ea4786742f1142210f1750f236d206b6585f982a4676c1fd6a09725ca", "generation_seconds_batch": 15.164196260971949, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3397960662841797, "target_z": 1.546874856330807, "ce": 6.0305562019348145, "genuineness": 4.068024158477783, "identity": 0.5844898819923401, "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_NEWS_reg/3197768b8cad08cd4a99c467cd9cac9223b1577e849611a5fa759fc25eda78ed.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "323138c96605b33fdba6442f5bf4fe747fa4374eac1520592bf178168c9f4f80", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "eb94328c7bed54c0cc0c80ee610bdef17462ce1989851b355556ac5d47405315", "generation_seconds_batch": 20.33513457106892, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.133810997009277, "genuineness": 4.086852550506592, "identity": 0.5752281546592712, "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_NEWS_reg/323138c96605b33fdba6442f5bf4fe747fa4374eac1520592bf178168c9f4f80.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "3267993a9d3e9b60fa41597e1eaa577d145e1891cbf4c026467456e2998256fe", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e9d424aad504c1bfe521ac33e88e4ae3842ccbb623e256a2474b8d5051d6003f", "generation_seconds_batch": 15.65497026196681, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1073410511016846, "target_z": 1.1250001216962575, "ce": 6.141753673553467, "genuineness": 4.093128204345703, "identity": 0.4923800826072693, "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_NEWS_reg/3267993a9d3e9b60fa41597e1eaa577d145e1891cbf4c026467456e2998256fe.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "32842f8703cec1023b83775d4e12b4d791d4c0d408f29a87903252ec9ab7423a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0cf7b29eaa0ce2107c0edc42c6e27f9482a6b4bfb4e438814f945ae65c708495", "generation_seconds_batch": 19.948915104032494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2450921535491943, "target_z": 1.374999932390968, "ce": 6.23706579208374, "genuineness": 4.507347106933594, "identity": 0.7176409363746643, "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_NEWS_reg/32842f8703cec1023b83775d4e12b4d791d4c0d408f29a87903252ec9ab7423a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "32b3e10a1f3910e2ca99876663b7eeb95586a90507e27b74fad3179145682659", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "bd683c3cf8514d29c787870b1faff1b9deba7716720d5413a8aef531be3bd719", "generation_seconds_batch": 21.20744886796456, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6940873861312866, "target_z": 0.3750000405654192, "ce": 6.4912309646606445, "genuineness": 4.306513786315918, "identity": 0.29985737800598145, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/32b3e10a1f3910e2ca99876663b7eeb95586a90507e27b74fad3179145682659.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "332aaae54921f6f1084d72a5e19c7f9cb817d6e3450bc9dda3e8cd084fedaffd", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d569b868be5a63152b71844c84f03900707895dbb6de88155769ee823aa2b0d3", "generation_seconds_batch": 15.893671412952244, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9437614679336548, "target_z": 0.8281250760601611, "ce": 6.038498878479004, "genuineness": 4.13078498840332, "identity": 0.8020766973495483, "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_NEWS_reg/332aaae54921f6f1084d72a5e19c7f9cb817d6e3450bc9dda3e8cd084fedaffd.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": "33b8325a63237466701325d46b3c08dc1d11e0d53797763b8ec838bd201c5362", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "84f8443f72e8505bf23587d0e86270456d363b43d59ad259d23d8f7a680c8b8e", "generation_seconds_batch": 19.701628713984974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.1894097328186035, "genuineness": 4.030367851257324, "identity": 0.7446576356887817, "hypothesis": "Quelqu'un a déplacé la petite chaise bleue. Elle était ici, près la grande bibliothèque. Nous la remettrons après le déjeuner.", "wer": 0.047619047619047616, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/33b8325a63237466701325d46b3c08dc1d11e0d53797763b8ec838bd201c5362.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "acd246fe45f3a7997f463fb04486c5d4bd8b9308d40a6c7e31843195e80c953c", "generation_seconds_batch": 14.886502415058203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.29590028524398804, "target_z": -0.34765624366165326, "ce": 5.847874641418457, "genuineness": 4.306513786315918, "identity": 0.6826620101928711, "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_NEWS_reg/33f63538ac2c49676cc53462aa2024b47e58cb29b8690774a6580785ecb0e60e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "3428ccf909af967afd3c25a01603d1219ebca910e910f5b45ef4e6b4180269e0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a0a60f07241b78070a1383ab31d4a915b6583d6782e4bf64f48b0e30edebe3cc", "generation_seconds_batch": 19.909621835919097, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.140475034713745, "target_z": 3.0, "ce": 6.419747352600098, "genuineness": 4.18099308013916, "identity": 0.6261212229728699, "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_NEWS_reg/3428ccf909af967afd3c25a01603d1219ebca910e910f5b45ef4e6b4180269e0.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": "3446344a16a76ae15b67d641b4a6154e07b8637a80cad9a4e096f95d0367d1f6", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The keys are on the kitchen table.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "aa71f0460741c79cb7b38550eff7bd1235c7cc2557eaed51930a8be24ee14bb1", "generation_seconds_batch": 14.663563595036976, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6325175762176514, "target_z": 2.078125211278225, "ce": 6.316492557525635, "genuineness": 3.5973212718963623, "identity": 0.36402955651283264, "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_NEWS_reg/3446344a16a76ae15b67d641b4a6154e07b8637a80cad9a4e096f95d0367d1f6.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "3c8b61195cd3682f4e69876435e5ab2aa5437a9e540358984af2e65bf58f75a0", "generation_seconds_batch": 13.991360471001826, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.913628339767456, "target_z": 0.7734374822526291, "ce": 6.356205940246582, "genuineness": 3.948779582977295, "identity": 0.457186758518219, "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_NEWS_reg/347b084eb1c87968a3fc14b855043a2de140e81cf0874b7d65525285bfb31414.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "3492f03347426838fb62f0f084991a5fbc6356ff99d4a37d2efd6c487f2c9298", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "dc27547da38e3626607c216c4c0b1e84286a7469719a8eab51f316de1a4abeed", "generation_seconds_batch": 15.57495532091707, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.221180438995361, "genuineness": 4.18099308013916, "identity": 0.5778810977935791, "hypothesis": "La puerta sigue abierta. He dejado tu abrigo sobre la silla. Dime cuándo estás listo. ¿En qué entendido?", "wer": 0.2, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/3492f03347426838fb62f0f084991a5fbc6356ff99d4a37d2efd6c487f2c9298.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "34d0b7e81a7e34ab644315a5f15963863d7a507f9a1f96edfb8593f769c11d19", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "406d56f5ec03895e5d4ddc3cfe08a3eb18ad2c6bb8051e726513f4a4f6c75b7b", "generation_seconds_batch": 18.767864911002107, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7005444765090942, "target_z": 0.38671879521354013, "ce": 6.41180419921875, "genuineness": 4.20609712600708, "identity": 0.14297211170196533, "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_NEWS_reg/34d0b7e81a7e34ab644315a5f15963863d7a507f9a1f96edfb8593f769c11d19.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": "359ff98b4050287f9699cbfd545ebc7737922f6741cd672dae19ebdce9d7d325", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Escribí el número en el sobre.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9d1504e1c30f25203e1e62150fa9b391cac56b0aa1c1094ddd75a8ca7dc43920", "generation_seconds_batch": 14.610306482063606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.054380416870117, "target_z": 2.8437497937924525, "ce": 6.308549880981445, "genuineness": 4.042920112609863, "identity": 0.5491284132003784, "hypothesis": "Escribí el número en el sobre, lélelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/359ff98b4050287f9699cbfd545ebc7737922f6741cd672dae19ebdce9d7d325.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "35c27e99b0b1ec5ffa584a556d039d67cc2141e5ae8283c54704117f3f4cae43", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "8a63f67b254ad9fafb02d64959bb03c46ff5d1e212187945605f24bd07498dfe", "generation_seconds_batch": 19.33410882903263, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1546928882598877, "target_z": 1.2109373673172747, "ce": 6.09409761428833, "genuineness": 4.519899368286133, "identity": 0.5595442056655884, "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_NEWS_reg/35c27e99b0b1ec5ffa584a556d039d67cc2141e5ae8283c54704117f3f4cae43.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "3681d60e35a5f00ca3fd7d83e3f72cb0cdaa572669176fcb727d9daa57d7e2a0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "41df4e550c6a0fa0ea18dfed20cff59f97e3ce277ca261344bd98cc1457db209", "generation_seconds_batch": 21.557112255948596, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4331963062286377, "target_z": 3.531249922249613, "ce": 6.308549880981445, "genuineness": 3.9425034523010254, "identity": 0.6936581134796143, "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_NEWS_reg/3681d60e35a5f00ca3fd7d83e3f72cb0cdaa572669176fcb727d9daa57d7e2a0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "368ef4e9803a6944fc467872c4ce58ee6fe8d0b5aacda3ee40fb60e68ad40eb3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9e7d5a8e2072090a007f8bbdd94a42156f8f54b9b5f05a495402836a1105ab0b", "generation_seconds_batch": 18.766539247939363, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5595647096633911, "target_z": 0.13085939084586687, "ce": 6.427689552307129, "genuineness": 4.344170093536377, "identity": 0.5059025883674622, "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_NEWS_reg/368ef4e9803a6944fc467872c4ce58ee6fe8d0b5aacda3ee40fb60e68ad40eb3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "376169c1b3fab781c3ee3254b8de9f2a101d44ced0d09546bc2b53c1a0bb3555", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ddbf1b3eace473feedf23e190a06ffa641481904917df4efa891564a0a02e520", "generation_seconds_batch": 15.915263572940603, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.417281150817871, "target_z": 1.6874999121082583, "ce": 6.268836498260498, "genuineness": 4.18099308013916, "identity": 0.7294063568115234, "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_NEWS_reg/376169c1b3fab781c3ee3254b8de9f2a101d44ced0d09546bc2b53c1a0bb3555.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "381a557e614519b5f49f7c3d4f92ca47cf164cda287e8f21baa422bd211b625c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "1ef7cb252ae9f250281529213ba5518fdc4fdecc0c1237355ff99cf26fbe9bab", "generation_seconds_batch": 21.359934424981475, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7100024223327637, "target_z": 2.2187498343578715, "ce": 6.09409761428833, "genuineness": 3.9550557136535645, "identity": 0.6536681652069092, "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_NEWS_reg/381a557e614519b5f49f7c3d4f92ca47cf164cda287e8f21baa422bd211b625c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "383f34cd412598759b24e13faa06d31943615a20133922860d8ad2d7ab87075b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0d4e17e67ea8ff57092f42586e5925bdf4244ecb76464cb3fa531311400f66e8", "generation_seconds_batch": 14.405578007921576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3397960662841797, "target_z": 1.546874856330807, "ce": 6.165581703186035, "genuineness": 4.243753433227539, "identity": 0.5658204555511475, "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_NEWS_reg/383f34cd412598759b24e13faa06d31943615a20133922860d8ad2d7ab87075b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "38a7ff45ad5e8314c573880735ad38c2ede4f99893e3505bfb0b0bf030d624e1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b2524f2b622b6087eb76d7cb783bc3226a317da4d4cd5f22459d14c95c0774bd", "generation_seconds_batch": 19.21775805298239, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5980796813964844, "target_z": 2.015625042255645, "ce": 5.927301406860352, "genuineness": 3.9927117824554443, "identity": 0.560361921787262, "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_NEWS_reg/38a7ff45ad5e8314c573880735ad38c2ede4f99893e3505bfb0b0bf030d624e1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "38daada8c245ac569aeac4909f481d11b90f7f01a6948ab6d510c0a1f8d365f5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2709ed999ef259e9693261b9a00852bbff8f6432bb58ea50c3de6544bd2e963a", "generation_seconds_batch": 14.772921173018403, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5714026689529419, "target_z": 0.15234370225112115, "ce": 6.38797664642334, "genuineness": 4.469690799713135, "identity": 0.40588539838790894, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/38daada8c245ac569aeac4909f481d11b90f7f01a6948ab6d510c0a1f8d365f5.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": "392b68644623fdc05a6051612cc2a9f8d82964492e878f09c12f9bc56d6cb9ca", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6f2cfeef790ee7b97a926e8f96cda2a468eb6eafdbccfc8f4ec6ca5a05d8c620", "generation_seconds_batch": 14.233767705969512, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0427701473236084, "target_z": 1.0078125752150482, "ce": 6.38003396987915, "genuineness": 4.13078498840332, "identity": 0.4720049500465393, "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_NEWS_reg/392b68644623fdc05a6051612cc2a9f8d82964492e878f09c12f9bc56d6cb9ca.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "3950f8eaaedc996bcd3c004fcf56c6f8b8b8d01b84acca6bccb785e12ece654b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "291509ad9971360940c1595c81f60e08e5db45ba9b57b8a14aa9d5507dfa45ed", "generation_seconds_batch": 18.833007132983766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0027239322662354, "target_z": 2.7500001893052897, "ce": 6.181467056274414, "genuineness": 4.168440818786621, "identity": 0.4181820750236511, "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_NEWS_reg/3950f8eaaedc996bcd3c004fcf56c6f8b8b8d01b84acca6bccb785e12ece654b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "3960942167f552796eab00d2453292596d563b2a603ab128cc3b4c1c556bb7dc", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "aa01f47daca1dae7955d1837135fecbf26a12ca95960d5fc6f611969598e0abc", "generation_seconds_batch": 15.576549438992515, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0363130569458008, "target_z": 0.9960938205669272, "ce": 6.300607204437256, "genuineness": 4.68307638168335, "identity": 0.7543286681175232, "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_NEWS_reg/3960942167f552796eab00d2453292596d563b2a603ab128cc3b4c1c556bb7dc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "3971d14af063aff3762a4e1af4e3fb9f2e5b710a911ac4308cd9ad5ba447ec52", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9da7bd1886360bb1ebe88f3b5ce6dc98fd65a5dd8dae1ab2cd8f595a4b9b168c", "generation_seconds_batch": 15.295344428042881, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6755647659301758, "target_z": 2.1562500980330963, "ce": 6.268836498260498, "genuineness": 4.218649387359619, "identity": 0.5712406635284424, "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_NEWS_reg/3971d14af063aff3762a4e1af4e3fb9f2e5b710a911ac4308cd9ad5ba447ec52.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "3a05a5b8bb0092a60f749bdf67a248487639011ede5f6a4f07e3ef9bb26c525b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c81d31ef4d1760b0712bfb4c1922499accd8cee99be1a97eaaeefeef5eb34986", "generation_seconds_batch": 19.453250244958326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8834952712059021, "target_z": 0.7187499966195484, "ce": 6.260893821716309, "genuineness": 4.055472373962402, "identity": 0.5716791152954102, "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_NEWS_reg/3a05a5b8bb0092a60f749bdf67a248487639011ede5f6a4f07e3ef9bb26c525b.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": "3a2e0311769d08845ec31d0c044165facb6adbd300a82b7533a64c5afc752ba6", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "5102b23936b5ca1873c3518401c9b1c47eadf102d75219df312192c8d81d2642", "generation_seconds_batch": 20.62028798495885, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 6.109982967376709, "genuineness": 3.961331605911255, "identity": 0.5636101961135864, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/3a2e0311769d08845ec31d0c044165facb6adbd300a82b7533a64c5afc752ba6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "3a98a536cbe6805158ba8dc400642cbb291a52abe85d000da7f9132eead4cbf7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ded60ce804de849b643738a07e723212043b1f1e31727f6b6f9552444a392219", "generation_seconds_batch": 19.894581312895752, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9510672092437744, "target_z": 2.656250152120322, "ce": 6.316492557525635, "genuineness": 3.9048471450805664, "identity": 0.5105301141738892, "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_NEWS_reg/3a98a536cbe6805158ba8dc400642cbb291a52abe85d000da7f9132eead4cbf7.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": "3abf36f4b8280f1301ce9bc4ad06cf2694abb5ab59c5c8e3c4287ec4b529253f", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "8d5bf176f8483418d169ddaca40fa11c904d5801bafb047c6e955d2d7136eb6c", "generation_seconds_batch": 17.38785416900646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7702686786651611, "target_z": 2.3281250219729355, "ce": 6.24500846862793, "genuineness": 4.3504462242126465, "identity": 0.5726689696311951, "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_NEWS_reg/3abf36f4b8280f1301ce9bc4ad06cf2694abb5ab59c5c8e3c4287ec4b529253f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "3b1a91a6e1dad547b40274436f7e1ae1fd4f2a595eb255cc715e0ca6738a1b31", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6b8d965f4432d91a34ab6fa28ec13a3af7f0eb460c26e4b8f29cdb14d5563efa", "generation_seconds_batch": 19.06925861700438, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3311867713928223, "target_z": 1.5312501385985156, "ce": 6.268836498260498, "genuineness": 4.519899368286133, "identity": 0.590912938117981, "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_NEWS_reg/3b1a91a6e1dad547b40274436f7e1ae1fd4f2a595eb255cc715e0ca6738a1b31.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "3b6ed13351fee40840ef44e1d853f23be138afb5490343bd7695460bd4d5da0d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "367daf24af7b23027335890017433a093f274dddf5d54ab691a72ebf121954db", "generation_seconds_batch": 20.091951805981807, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6411268711090088, "target_z": 2.0937499290105164, "ce": 6.24500846862793, "genuineness": 4.325342178344727, "identity": 0.6622202396392822, "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_NEWS_reg/3b6ed13351fee40840ef44e1d853f23be138afb5490343bd7695460bd4d5da0d.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "376576757c409b4b4025a2bb2b535b029482604f9e5dcfa20306405f61d81e91", "generation_seconds_batch": 14.189571517985314, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7616591453552246, "target_z": 2.312499871542839, "ce": 6.149696350097656, "genuineness": 4.268857479095459, "identity": 0.7293548583984375, "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_NEWS_reg/3c590f9302048ad206a88d3c2440f1028cc9eea1f1662250baf365be21a6ebfa.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "3c8b0faa8a70b886b2f1169c791bcf09b3703b027629e8f3413ec5a4abebd679", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "35d88a37a1dcad7970f0cef9c594e6bd8b29cd59412da84a43b5b32bc0c9d968", "generation_seconds_batch": 19.38454408699181, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8834952712059021, "target_z": 0.7187499966195484, "ce": 6.038498878479004, "genuineness": 4.381826400756836, "identity": 0.6398540735244751, "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_NEWS_reg/3c8b0faa8a70b886b2f1169c791bcf09b3703b027629e8f3413ec5a4abebd679.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "3cd58f1040d12a45c767f996af8f3914e6c13354dc3d0f25a8591ee0bb761480", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c4b562499c0aacc3009db54b866292554e841e2cd77c58ce8e596e0222fe33ba", "generation_seconds_batch": 15.118976655998267, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3139677047729492, "target_z": 1.4999998377383232, "ce": 6.173524379730225, "genuineness": 4.494794845581055, "identity": 0.6712287068367004, "hypothesis": "Je trouvai la petite boîte. Elle était près de la vieille fenêtre. Nous pouvons regarder ensemble ce qu'elle contient.", "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_NEWS_reg/3cd58f1040d12a45c767f996af8f3914e6c13354dc3d0f25a8591ee0bb761480.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": "3cf517ea692c776564b291e542d343b13baefa01ad21264df0dcccf117d62a00", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dbc6f92d9529a6a515a58152d1d194a0b4a8c712f2c5e60c755d1f003f416bba", "generation_seconds_batch": 15.518799489014782, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.545119285583496, "target_z": 3.7343751470496445, "ce": 6.1894097328186035, "genuineness": 3.9173994064331055, "identity": 0.7547246217727661, "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_NEWS_reg/3cf517ea692c776564b291e542d343b13baefa01ad21264df0dcccf117d62a00.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "3d0c66fefd5289099e0701abcf5c9a2aa3c237ed4cbee03c85c67c60ac4009ae", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "faccbbaa177567e5ebcc532b6910feb53b992b893911ed27c291563e5a8a2327", "generation_seconds_batch": 19.453250244958326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0599889755249023, "target_z": 1.0390624433774358, "ce": 6.24500846862793, "genuineness": 3.7918784618377686, "identity": 0.33536607027053833, "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_NEWS_reg/3d0c66fefd5289099e0701abcf5c9a2aa3c237ed4cbee03c85c67c60ac4009ae.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "3d113457ee44b71e4844cfbad93e25366d53ec592824c9378fde5e070e01dde2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "358e1aec50be888196089e5b94fec532afabe3c0eb78d234a87fb05a61b0e90c", "generation_seconds_batch": 19.76222982001491, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9510672092437744, "target_z": 2.656250152120322, "ce": 6.443574905395508, "genuineness": 4.193545341491699, "identity": 0.3213874101638794, "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_NEWS_reg/3d113457ee44b71e4844cfbad93e25366d53ec592824c9378fde5e070e01dde2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "3d34c678a44d6a0002bbd0b5f2e67ad8be81de34a20c4c6cf94b59c1118302be", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "30c784141ea7624a49fedf759dbe871555a4802068058d30276c04feb9ed3e0f", "generation_seconds_batch": 18.85329126601573, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1245598793029785, "target_z": 1.1562499898586451, "ce": 6.157639026641846, "genuineness": 4.381826400756836, "identity": 0.4459105134010315, "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_NEWS_reg/3d34c678a44d6a0002bbd0b5f2e67ad8be81de34a20c4c6cf94b59c1118302be.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "3e4b20b65d9a1b209dfa853a55b1ab0f53de64370554faac8eae64447f38351f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "9341702e123b73a000ef44f55f4f0f9b742f452476c9445eae534fbb6aae8024", "generation_seconds_batch": 18.844022584031336, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4000623226165771, "target_z": 1.6562500439458707, "ce": 6.332377910614014, "genuineness": 4.532451152801514, "identity": 0.7799655199050903, "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_NEWS_reg/3e4b20b65d9a1b209dfa853a55b1ab0f53de64370554faac8eae64447f38351f.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": "3e5d2f9e512f9da1f6f3f5715866461a22b92ff8301804b479b82f52c8cf1bcd", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Your message arrived after I had left.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "73f39aa62934e589fecd25dc4ea50da19ca17ceddcf5fff4b94f38d5c532b1ca", "generation_seconds_batch": 13.917710524052382, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0599889755249023, "target_z": 1.0390624433774358, "ce": 6.348263263702393, "genuineness": 3.8295347690582275, "identity": 0.34194687008857727, "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_NEWS_reg/3e5d2f9e512f9da1f6f3f5715866461a22b92ff8301804b479b82f52c8cf1bcd.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": "3f2815ca03e361686c58269b85219d366b5c12c1229339bb83bbeb91a2655980", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "ba4441a36e7fbcfb61e38acb1d889f87dc411ac6b2fa59cafded5927aa21dab4", "generation_seconds_batch": 19.107649150071666, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8477537631988525, "target_z": 2.468750077750387, "ce": 6.181467056274414, "genuineness": 3.8671910762786865, "identity": 0.684158980846405, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/3f2815ca03e361686c58269b85219d366b5c12c1229339bb83bbeb91a2655980.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "3fe5abe765d575dabf5620312868561d1701b586be358119c7eda45ea95f1377", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7af2919576280b33443697089806a1c29e7339d6ed62f0149e2c2d9b400dadc6", "generation_seconds_batch": 19.80228428193368, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5679466724395752, "target_z": 1.9609376647970156, "ce": 5.895530700683594, "genuineness": 3.7165660858154297, "identity": 0.42727330327033997, "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_NEWS_reg/3fe5abe765d575dabf5620312868561d1701b586be358119c7eda45ea95f1377.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "40344904bdd3b1782b7db04e53326e331c1be1e926a60c2121b988eb8c6b5edd", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fc0af6ad59982ce82dec1afae0ec7638f1d7abc0575f535031c21d14ade222a4", "generation_seconds_batch": 15.337763014016673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8047065734863281, "target_z": 2.3906251909955154, "ce": 6.24500846862793, "genuineness": 4.15588903427124, "identity": 0.6545645594596863, "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_NEWS_reg/40344904bdd3b1782b7db04e53326e331c1be1e926a60c2121b988eb8c6b5edd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "40994ccc7fe8b441ceb204d47e0a4980112444324709bdec3229bd368f2da709", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "f4a280b26c19e279fdf3db9a74188aac72abbeb2a7fc4abb0fafa045a03b0d5c", "generation_seconds_batch": 18.68280913599301, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1805212497711182, "target_z": 1.2578123859097585, "ce": 6.356205940246582, "genuineness": 4.306513786315918, "identity": 0.47225314378738403, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/40994ccc7fe8b441ceb204d47e0a4980112444324709bdec3229bd368f2da709.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "40ab9ff1f0dee5287b36fa362c2ebd57ac468158e408058935f9d7af5e9d304a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "34be807985f771034f35bcd7e93d52ce391d9efbd9641bf86ab800bf0d38742b", "generation_seconds_batch": 15.40027757990174, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3570151329040527, "target_z": 1.5781251571909993, "ce": 6.197352409362793, "genuineness": 4.3127899169921875, "identity": 0.7341156005859375, "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_NEWS_reg/40ab9ff1f0dee5287b36fa362c2ebd57ac468158e408058935f9d7af5e9d304a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "40ba990d72003e026859ee27a3e4db038863aaa7d093abb5697c6db325d322ff", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "850bda5907bd97a95b047dfd53e84410f4a95522204f7d0ce3b2a5f1cd3405f4", "generation_seconds_batch": 20.937172919046134, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8133158683776855, "target_z": 2.4062499087278066, "ce": 6.173524379730225, "genuineness": 3.986435890197754, "identity": 0.5888800621032715, "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_NEWS_reg/40ba990d72003e026859ee27a3e4db038863aaa7d093abb5697c6db325d322ff.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "41226539e354762be1c5892b073bb7b5c8d00aead705a83f1184f4d678179cfa", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ee3adf57b234d265bb7507e62de3230ff625b91ea67729f25624c0e7500f0969", "generation_seconds_batch": 19.74287048506085, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5248992443084717, "target_z": 1.8828123453443393, "ce": 6.38003396987915, "genuineness": 4.262581825256348, "identity": 0.29051244258880615, "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_NEWS_reg/41226539e354762be1c5892b073bb7b5c8d00aead705a83f1184f4d678179cfa.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "41a4dd84347723aed8b8f8d6f4b7f9456d60ac124040886291e680455097389b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "98785f0c803e6cb941540b360efe32ac302038233078370e7f8dc9bfdc575129", "generation_seconds_batch": 15.118976655998267, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2063496112823486, "target_z": 1.3046874045022423, "ce": 6.205295085906982, "genuineness": 4.344170093536377, "identity": 0.694710910320282, "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_NEWS_reg/41a4dd84347723aed8b8f8d6f4b7f9456d60ac124040886291e680455097389b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "41fb2a80270033868c10ee65281b181f2e4b2ee93daef9961567695304172fff", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4fc4cbaf6cad9c7ddf98501ef6c1c32a8d96798d0467eda5f57675958e13734a", "generation_seconds_batch": 20.05922737705987, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6927835941314697, "target_z": 2.187499966195484, "ce": 6.229123115539551, "genuineness": 3.948779582977295, "identity": 0.3539363741874695, "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_NEWS_reg/41fb2a80270033868c10ee65281b181f2e4b2ee93daef9961567695304172fff.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "42020d06a24ec1cea4bf28ed6527616749724dcc44d956774a8399df4217b9f6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cf812ad0f264904aeb9f34f52f9610c5e10a8008e74477f4a8e6cebe5c82015f", "generation_seconds_batch": 20.212513353093527, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6411268711090088, "target_z": 2.0937499290105164, "ce": 6.205295085906982, "genuineness": 4.375550270080566, "identity": 0.6776458621025085, "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_NEWS_reg/42020d06a24ec1cea4bf28ed6527616749724dcc44d956774a8399df4217b9f6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "42aafad2fc99b39dc632ac73406b4d4bcd906e3160894a49a36da1d513b247f2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5795f368f3f3bcca1f31512cc7729b7ea61d15e18732b9f1c2f9e83ca5fa30a2", "generation_seconds_batch": 21.407615242060274, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 5.919358730316162, "genuineness": 3.8797430992126465, "identity": 0.6748151779174805, "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_NEWS_reg/42aafad2fc99b39dc632ac73406b4d4bcd906e3160894a49a36da1d513b247f2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "42db862ffbb5dcc01dd78c3aa2b7e7ee3e526649781218aea6ba64691ecb7082", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5bbf55d5238e93e6425a6ed5d7057e311f0bf6167348b2878d15b154c04aeceb", "generation_seconds_batch": 15.915263572940603, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.451719045639038, "target_z": 1.7500000811308385, "ce": 6.173524379730225, "genuineness": 4.231201171875, "identity": 0.7782873511314392, "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_NEWS_reg/42db862ffbb5dcc01dd78c3aa2b7e7ee3e526649781218aea6ba64691ecb7082.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "43462bdfb647b6219cb0a5c6b1e96ae963d7796124dcafa8e5f753462a697d21", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eda8436d2594aa9606ec1e01728e9085b800d8a0c94454079aab6c2d679cf6c0", "generation_seconds_batch": 14.729625968029723, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 5.95112943649292, "genuineness": 4.055472373962402, "identity": 0.5773510336875916, "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_NEWS_reg/43462bdfb647b6219cb0a5c6b1e96ae963d7796124dcafa8e5f753462a697d21.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "f19d7ecaddf70fe60d3b82e26aa91d4832e06542693df5d80c927b4c26ba469d", "generation_seconds_batch": 14.886502415058203, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.1624538004398346, "target_z": -0.5898437495774436, "ce": 6.070269584655762, "genuineness": 4.582659721374512, "identity": 0.7151672840118408, "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_NEWS_reg/434f953004ee3838b31e4f505ba0cb9202b34ab37936979fff50edb64a6a996f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "435dbe6f7d83e6e8ae8d3465d63ee52ca07b528c96fc46570e42b633d3fbaa08", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5d081d6d62133972058d8c7c24fc808c3b7ba15876d88fb16fdeaff65bb37a6f", "generation_seconds_batch": 18.951040953979827, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.451719045639038, "target_z": 1.7500000811308385, "ce": 6.260893821716309, "genuineness": 4.137060642242432, "identity": 0.42432263493537903, "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_NEWS_reg/435dbe6f7d83e6e8ae8d3465d63ee52ca07b528c96fc46570e42b633d3fbaa08.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "43653e65b2e85d79586f86cfcf2163cf42a33312cc13e6e62147f391bc59b839", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "c460b34ad1e1179d552ed66e8ff9b2ef6d690ef22217f1ecc92a4f7c4c221b3f", "generation_seconds_batch": 20.1207884530304, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7823342084884644, "target_z": 0.5351562098571373, "ce": 6.3641486167907715, "genuineness": 4.030367851257324, "identity": 0.512766420841217, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/43653e65b2e85d79586f86cfcf2163cf42a33312cc13e6e62147f391bc59b839.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "43ac6a5665b44914b38924875a431b2c5c7fe4f0450f90e62c6f73d1cb67a39b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a5c46edc33c969c532bcd03fdffb95b6425272494bdc05032bef7234ac51ac9c", "generation_seconds_batch": 14.389843570068479, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 6.332377910614014, "genuineness": 3.8671910762786865, "identity": 0.3359084129333496, "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_NEWS_reg/43ac6a5665b44914b38924875a431b2c5c7fe4f0450f90e62c6f73d1cb67a39b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "43d2e4ddcb6a0b7d20026cfde1afbe0adf2add8d3830cd63ce662489102030a3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ba1344a5e1e6417e52214431be0bcfc43e27e7116efc04923ca5528aa0aeda76", "generation_seconds_batch": 19.963497626013122, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7285251617431641, "target_z": 0.4374999932390968, "ce": 6.205295085906982, "genuineness": 3.8671910762786865, "identity": 0.5625320672988892, "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_NEWS_reg/43d2e4ddcb6a0b7d20026cfde1afbe0adf2add8d3830cd63ce662489102030a3.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "198faf860f391bb8464046d5c9bc1c4f2d6117955c23683ff3c0f67f63f6d7bf", "generation_seconds_batch": 15.394858321989886, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1460835933685303, "target_z": 1.1953126495849833, "ce": 6.213237762451172, "genuineness": 4.425758361816406, "identity": 0.7050122618675232, "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_NEWS_reg/43d771a5c055e2aea1186dc99f4cdef2b4c6bc8c6b2779ad35a29ca11c75a91c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "43dc5a2ab5c94966ed89925ffe2fe6c2fcd61e2ec4381cb7d854db4804b01535", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "39868e0d2cc2b5830cc2538c3774469e3f4b49d1cf9bf61b6bc97de9ea950b89", "generation_seconds_batch": 15.899023808073252, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3139677047729492, "target_z": 1.4999998377383232, "ce": 5.998785495758057, "genuineness": 4.419482707977295, "identity": 0.45920366048812866, "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_NEWS_reg/43dc5a2ab5c94966ed89925ffe2fe6c2fcd61e2ec4381cb7d854db4804b01535.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "43ed2ce9b0e3fa1aa5391cd64747682d20f5f9a027b304df8c19fd7ce7fccb39", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "70d9d45118c5e0a29e76842ed28cbf9a785ea019645d6d26acc3399fb80e9184", "generation_seconds_batch": 19.831851085997187, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1891307830810547, "target_z": 1.2734375363398547, "ce": 6.046441555023193, "genuineness": 4.457139015197754, "identity": 0.8272941708564758, "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_NEWS_reg/43ed2ce9b0e3fa1aa5391cd64747682d20f5f9a027b304df8c19fd7ce7fccb39.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "4428a89d80de09526c83bfb4772435507e8baf3eaa4174b5f4add9293ffa8c86", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "69bef7f87ff7f45ce8e585263646aa990d80623fe9f62033612b7f6f521184a9", "generation_seconds_batch": 15.750117280054837, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.890800952911377, "target_z": 2.546874964505258, "ce": 6.173524379730225, "genuineness": 4.306513786315918, "identity": 0.6362017393112183, "hypothesis": "Ik heb het kleine doosje gevonden, het stond naast het oude 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_NEWS_reg/4428a89d80de09526c83bfb4772435507e8baf3eaa4174b5f4add9293ffa8c86.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "4429642a6ae07bea21c2dc455ab2bab96202cc285cc25a5b41a59638631e7c1c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "46f9f1371d802b8c87a196159a8c0c09533ff83ceaecbecff4c5a92190c3ae42", "generation_seconds_batch": 19.86863025999628, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.090122103691101, "target_z": 1.0937500371849675, "ce": 6.292664527893066, "genuineness": 4.457139015197754, "identity": 0.6178077459335327, "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_NEWS_reg/4429642a6ae07bea21c2dc455ab2bab96202cc285cc25a5b41a59638631e7c1c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "444c7ad8c3368cdbcf365f3e19862b2b1ecedd71cdf506083e000f726c2b47da", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "6e2402c5663037cd9d49ff2bf9e8d5798fda92129fb491f647d78ee71285cec9", "generation_seconds_batch": 21.27678560791537, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.047074794769287, "target_z": 1.0156249340811938, "ce": 6.475345611572266, "genuineness": 4.319066047668457, "identity": 0.3554721474647522, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/444c7ad8c3368cdbcf365f3e19862b2b1ecedd71cdf506083e000f726c2b47da.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "44bdab0b9a770b6b342aea606ed3da42a1b626a958e30cc79f60f5e9831a1b68", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "101fd7423956906bea107f5bc29b7f296769589f3caf12655bf7418a6abe1f77", "generation_seconds_batch": 19.005076337023638, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2795300483703613, "target_z": 1.437500101413548, "ce": 6.038498878479004, "genuineness": 4.331617832183838, "identity": 0.7988333702087402, "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_NEWS_reg/44bdab0b9a770b6b342aea606ed3da42a1b626a958e30cc79f60f5e9831a1b68.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": "44e9d963b63698189b75a4105b943a7edf2108a8682ef2533f8b6ba9203fa328", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The keys are on the kitchen table.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "704279be687079336825c6473961fb758378ee177833cf858313601d3c3b0956", "generation_seconds_batch": 22.14463098300621, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.284721851348877, "genuineness": 3.886019229888916, "identity": 0.3304606080055237, "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_NEWS_reg/44e9d963b63698189b75a4105b943a7edf2108a8682ef2533f8b6ba9203fa328.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": "4502fc72bedafd72b0d4699dab00cf78d06558e53309172b85e8f8d61df9a532", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "03a3014e43b7ae12748f656d6cd5285769e48585d17fb0875c2b2c0bef539d96", "generation_seconds_batch": 19.660428868955933, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.088818311691284, "target_z": 2.9062499628150325, "ce": 6.260893821716309, "genuineness": 3.9173994064331055, "identity": 0.581425130367279, "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_NEWS_reg/4502fc72bedafd72b0d4699dab00cf78d06558e53309172b85e8f8d61df9a532.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "450bb24d311012c4b622b99821d0438715393d4b0e637008e2da5d301d97fd23", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "37a9246f19788966ad94ad7b8d564ce1779e7b3c1141750e9d5413a423fa5d6f", "generation_seconds_batch": 19.524337145965546, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.016941785812378, "target_z": 0.9609375566225643, "ce": 6.141753673553467, "genuineness": 4.055472373962402, "identity": 0.3476276397705078, "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_NEWS_reg/450bb24d311012c4b622b99821d0438715393d4b0e637008e2da5d301d97fd23.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": "453a657a1d1725251b18af1c25c3424af2054bbbe23cd9461cb33d64dd0dfc0d", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2308b653cbf6812b1c2acfbbff35699b2952e5d385404b1e4a3020107a644e63", "generation_seconds_batch": 14.949785306002013, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.261007308959961, "target_z": 3.2187499425323227, "ce": 6.268836498260498, "genuineness": 3.9425034523010254, "identity": 0.5391021966934204, "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_NEWS_reg/453a657a1d1725251b18af1c25c3424af2054bbbe23cd9461cb33d64dd0dfc0d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "45432934e979e7249bd9214f2554b02b5c7528dcd3bab412886ad6b0c828a3b6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cf9399bab62c9563cd567b8ef4a76ac834ae0faf535a085b7f30e9604df5c90c", "generation_seconds_batch": 18.862212941981852, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.141753673553467, "genuineness": 3.8420867919921875, "identity": 0.40637749433517456, "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_NEWS_reg/45432934e979e7249bd9214f2554b02b5c7528dcd3bab412886ad6b0c828a3b6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "45c57e35d69cc6dc30538e79a6a83c6f9f9c6dad6329a75a46ee7f8ca87df9ec", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fca73a1db8b8b2a842b4c60ea002dd861a70f77fd7b031fc70669db46743ca27", "generation_seconds_batch": 19.06925861700438, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4000623226165771, "target_z": 1.6562500439458707, "ce": 6.141753673553467, "genuineness": 4.419482707977295, "identity": 0.7079924941062927, "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_NEWS_reg/45c57e35d69cc6dc30538e79a6a83c6f9f9c6dad6329a75a46ee7f8ca87df9ec.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "a8522e592854b760126301f9e7c63f2e39dd56f9cc08176b1b2a977ab7c0e979", "generation_seconds_batch": 14.41017354291398, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6497364044189453, "target_z": 2.1093750794406128, "ce": 6.300607204437256, "genuineness": 4.21237325668335, "identity": 0.5468757152557373, "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_NEWS_reg/45f9ec8497c8691f9e18d794cf98f47c522bfab041b36df5accabf5c5c456412.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "460deeb42b9c7d61b2e7303e24b4ee979b7b6976b81ee01f276ab550ae905c36", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "eada367b6a088137c73c0b6cbbecdd7d48028a673ccdaddf3e2a42de53357994", "generation_seconds_batch": 19.379194129025564, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 5.919358730316162, "genuineness": 3.8797430992126465, "identity": 0.465739905834198, "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_NEWS_reg/460deeb42b9c7d61b2e7303e24b4ee979b7b6976b81ee01f276ab550ae905c36.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "46c5755c8517b94a2a30039c864b9dae6b30e2972bd609187dfee65f5bb84265", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "eb2f92960fcf43dd1cdf22193c8757a7edee29e1d3cc7128696af8ba5b96e724", "generation_seconds_batch": 19.71063242899254, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8533622026443481, "target_z": 0.6640625109864677, "ce": 6.23706579208374, "genuineness": 3.8671910762786865, "identity": 0.18998974561691284, "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_NEWS_reg/46c5755c8517b94a2a30039c864b9dae6b30e2972bd609187dfee65f5bb84265.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": "47ac6b793de90b2796af5df379243fc3d1d36eab6a4ba41f76c8d83af58543b8", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cb3f16e0efe9d0299e30e0a554e1f71f54a617aafc59b4ca2078281f11d3bc97", "generation_seconds_batch": 19.98620509193279, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0974278450012207, "target_z": 2.9218751132451284, "ce": 6.213237762451172, "genuineness": 3.9676074981689453, "identity": 0.7276618480682373, "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_NEWS_reg/47ac6b793de90b2796af5df379243fc3d1d36eab6a4ba41f76c8d83af58543b8.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "a406a3d01720b0bb6e2586e5cab5d346777af653ffaa98e2579829e533d6f05a", "generation_seconds_batch": 13.922070267028175, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5636417865753174, "target_z": 1.953124873233065, "ce": 6.1020402908325195, "genuineness": 4.570107460021973, "identity": 0.8292563557624817, "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_NEWS_reg/47d00b404f769e82194cfb39f2be7a6bd3cc5bfaac4f38356b0fada2abdb9f4d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "480320ee3daaaa07d2de92a4766630cab184f0050fa0f83e973d005bc92a9387", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "21dfe4c7966c06909d6ae3048ea0364d3b298414cdd6fe22adf162c73810acb0", "generation_seconds_batch": 19.122632598038763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5119850635528564, "target_z": 1.8593748360480973, "ce": 5.990842819213867, "genuineness": 3.9425034523010254, "identity": 0.3234822750091553, "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_NEWS_reg/480320ee3daaaa07d2de92a4766630cab184f0050fa0f83e973d005bc92a9387.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "484a44190994d5844aad69d04b7529d3e65516d2f58b5bf900d920468787543c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2f39c9ba7d710332bb35a660d0c6271f0d482f92be069db2dd37cd8fbb63ef23", "generation_seconds_batch": 19.909621835919097, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.38003396987915, "genuineness": 4.262581825256348, "identity": 0.6596747636795044, "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_NEWS_reg/484a44190994d5844aad69d04b7529d3e65516d2f58b5bf900d920468787543c.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": "48633ba957220abb1fcdb010ea54671b6361618eeb9d1c59630337f95f003cb6", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9a8e08f68b43bb4ad257d60a2d4edbde3ee81ad543204fe2657a042294e715bc", "generation_seconds_batch": 14.014309831894934, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0715994834899902, "target_z": 2.875000094652645, "ce": 6.117925643920898, "genuineness": 3.986435890197754, "identity": 0.48055046796798706, "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_NEWS_reg/48633ba957220abb1fcdb010ea54671b6361618eeb9d1c59630337f95f003cb6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "48ca2cf01a9af5f0def3b4d199e2b697c1555019d0fa9c5d67c81c0ab8aa2cdc", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d8b92c690597c6567388f6c0b2ac24cfad51a1d582328aaa9feffc66c5ed0a14", "generation_seconds_batch": 19.765931225032546, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8017054796218872, "target_z": 0.5703124738015001, "ce": 6.260893821716309, "genuineness": 4.118232727050781, "identity": 0.46000945568084717, "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_NEWS_reg/48ca2cf01a9af5f0def3b4d199e2b697c1555019d0fa9c5d67c81c0ab8aa2cdc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "48e93160d806aa2a511bb278434497a484953334abfb696a2a7dedf25c82e5d0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "42f5d4f0c30acba28c4b0bb6f2e0f61593c8bf47ba04246ec89d8250362f5f71", "generation_seconds_batch": 19.61908814799972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1589977741241455, "target_z": 1.2187501588812253, "ce": 6.1894097328186035, "genuineness": 4.394378662109375, "identity": 0.6498537063598633, "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_NEWS_reg/48e93160d806aa2a511bb278434497a484953334abfb696a2a7dedf25c82e5d0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "49252714e67bf979a053387e9f7266d1e5e7edcc0736704033de3dd6185c5b11", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "11e7ec13c221d9fa5b644025eae292859099fd5698d3cee3f44e25e02d34524c", "generation_seconds_batch": 14.920606210944243, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9510672092437744, "target_z": 2.656250152120322, "ce": 6.38797664642334, "genuineness": 3.8797430992126465, "identity": 0.5648859739303589, "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_NEWS_reg/49252714e67bf979a053387e9f7266d1e5e7edcc0736704033de3dd6185c5b11.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "49979af0d125cbd083699409b14af1ebfc3fd6e4d4ac0d05f559144283da1124", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "311aca0bd8203f1aff94fdb6c618be416750e833c3dafe50868836cf7e92e0b7", "generation_seconds_batch": 15.45976946305018, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1245598793029785, "target_z": 1.1562499898586451, "ce": 6.268836498260498, "genuineness": 4.231201171875, "identity": 0.6448639631271362, "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_NEWS_reg/49979af0d125cbd083699409b14af1ebfc3fd6e4d4ac0d05f559144283da1124.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "e0dd2c0d529a5840efbb16eea7dd4041b5d5c95ec3f229653dd5cb378d297a4e", "generation_seconds_batch": 14.950288043939508, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8748857975006104, "target_z": 0.7031249543639034, "ce": 5.998785495758057, "genuineness": 4.356721878051758, "identity": 0.6166708469390869, "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_NEWS_reg/49d468ae6540cc7c1e5b0e6883c25e2b29599458c9de0a5398000eaca62677ab.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "49f36a896d3c7e318fb433516bc77335cde63862fd0bf14b784bce62dd6295c0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8e6affc352b7b15118e66e3b1969d289e1d4680ba68c7131f1fc2de7e476c86a", "generation_seconds_batch": 14.485948589979671, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.174912929534912, "target_z": 3.06250016902258, "ce": 6.38003396987915, "genuineness": 4.005263805389404, "identity": 0.6680704951286316, "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_NEWS_reg/49f36a896d3c7e318fb433516bc77335cde63862fd0bf14b784bce62dd6295c0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "4a5e438201de168bfe83f21cd2db3b510e0324290187c5346fbd605f93280e8b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cfeeb7cef9fdc084ced739e3a948f8c29710b675478bc95ffdbfca08be27f3eb", "generation_seconds_batch": 21.62259983702097, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.417281150817871, "target_z": 1.6874999121082583, "ce": 6.308549880981445, "genuineness": 4.457139015197754, "identity": 0.6988572478294373, "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_NEWS_reg/4a5e438201de168bfe83f21cd2db3b510e0324290187c5346fbd605f93280e8b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "4ae00288624d3c735d04833270f38acf60b92857b6266bebf4b431e354c66977", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8a089189eca5e892d90912a5ab0f558743eaddf044c9a24c5a1f8675a3defdac", "generation_seconds_batch": 19.107333330088295, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3656244277954102, "target_z": 1.5937498749232908, "ce": 6.284721851348877, "genuineness": 4.319066047668457, "identity": 0.5562736988067627, "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_NEWS_reg/4ae00288624d3c735d04833270f38acf60b92857b6266bebf4b431e354c66977.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "4b65e093145686d132de152a15760168f5fd228cd9b4793609193a46e4b5ab4a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bd05838514e27897295afea78e0fa87731b1336edacb9d21eaf90f594b5fb1a4", "generation_seconds_batch": 20.15223007905297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0642937421798706, "target_z": 1.0468750185924838, "ce": 5.974957466125488, "genuineness": 4.419482707977295, "identity": 0.672289252281189, "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_NEWS_reg/4b65e093145686d132de152a15760168f5fd228cd9b4793609193a46e4b5ab4a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "4c5b2a51b48e9f55c8e3293ef4d9c2ef7159f93710cbb1076e35247b2b410626", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7a09fc5b38d06591b1e3135f644167e3210003c3089aad5e442309852bd2e495", "generation_seconds_batch": 20.47050705202855, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8877999782562256, "target_z": 0.7265624636601453, "ce": 6.316492557525635, "genuineness": 4.143336772918701, "identity": 0.23469725251197815, "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_NEWS_reg/4c5b2a51b48e9f55c8e3293ef4d9c2ef7159f93710cbb1076e35247b2b410626.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": "4cb48157164ae0f5f6054968ba89aa9f6ffd577581a6c5d74c0e250b407dfbeb", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "df3065cc52ba9a10a917e9c878f04a07472c1c8a214735f3e773e1ef24e93c59", "generation_seconds_batch": 20.435180311091244, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3815395832061768, "target_z": 3.4374998850646454, "ce": 6.2767791748046875, "genuineness": 3.9676074981689453, "identity": 0.6482782363891602, "hypothesis": "Ik schreef het nummer op de envelop, lees het goed en controleer de laatste twee cijfers. Bel me wanneer je even tijd hebt.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/4cb48157164ae0f5f6054968ba89aa9f6ffd577581a6c5d74c0e250b407dfbeb.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "4cb9f75722e3515c00543df76b978c4091beee88cb448edbac21e361885de99e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "71eaac2fb57a69a30917940161be2f2968ad8befd1e0b5110125d900375b8d21", "generation_seconds_batch": 20.549167230026796, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6325175762176514, "target_z": 2.078125211278225, "ce": 6.054384231567383, "genuineness": 4.068024158477783, "identity": 0.6089466214179993, "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_NEWS_reg/4cb9f75722e3515c00543df76b978c4091beee88cb448edbac21e361885de99e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "4cfdd785f07f82178a7952d646b17d96834c5502bb6084761895f1ee7ae335ec", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9999dfab4c175d1abd07b840778baad3d0f94c4824594465cbd5679cf46764aa", "generation_seconds_batch": 20.522779140039347, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 6.165581703186035, "genuineness": 4.281409740447998, "identity": 0.7362903356552124, "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_NEWS_reg/4cfdd785f07f82178a7952d646b17d96834c5502bb6084761895f1ee7ae335ec.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "4d0efef21731c52e217b949309763803da6614d5c108d46039a0557af60e584a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "204a2780f920733c7f309db527cddc927f9e001c65c2374eef2317871f248c37", "generation_seconds_batch": 18.708634820999578, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.329882860183716, "target_z": 3.343749847879678, "ce": 6.41180419921875, "genuineness": 4.005263805389404, "identity": 0.42058098316192627, "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_NEWS_reg/4d0efef21731c52e217b949309763803da6614d5c108d46039a0557af60e584a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "4d30853cfba1002a96b7857608b2a10bae5b75d942cccf809a05d9d8577fa446", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "14da8ae2a308fd0708090e52e044c8be6cc774bd1fbca9e3c17d38cd263d5519", "generation_seconds_batch": 20.897332945023663, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8533622026443481, "target_z": 0.6640625109864677, "ce": 6.157639026641846, "genuineness": 3.9676074981689453, "identity": 0.2936311364173889, "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_NEWS_reg/4d30853cfba1002a96b7857608b2a10bae5b75d942cccf809a05d9d8577fa446.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "4d5c4b9f3506ad72cf68fd1cef6659d6a6021836d4911e8dcd308fefa2df2148", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fadb5c72e4c6e623ecc54583afadec67d8bd111f213965f7c9fc7b043c675f0f", "generation_seconds_batch": 19.338424205081537, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7788782119750977, "target_z": 2.3437501724030314, "ce": 6.229123115539551, "genuineness": 4.15588903427124, "identity": 0.7183480262756348, "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_NEWS_reg/4d5c4b9f3506ad72cf68fd1cef6659d6a6021836d4911e8dcd308fefa2df2148.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "4d92bf96a563a0ee1399f95ea0361af538fc43cd1f78637c64709278009b0056", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "72e78927ac9dc545a592c6f9dc41be68dbaebb685930841ca56ab53ab6a8a4f9", "generation_seconds_batch": 19.998955778079107, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1848260164260864, "target_z": 1.2656249611248065, "ce": 6.2767791748046875, "genuineness": 4.545003414154053, "identity": 0.7187977433204651, "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_NEWS_reg/4d92bf96a563a0ee1399f95ea0361af538fc43cd1f78637c64709278009b0056.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "4db15cd6f3d16f431c0d287d9092b4fd4b85b45586755c9019c40faa64e39c84", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3aa08bd56dcfea6505a5ca6966c530167fbf75c8b8426a2e6e9caf8356f68385", "generation_seconds_batch": 20.34802017000038, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3742339611053467, "target_z": 1.609375025353387, "ce": 6.141753673553467, "genuineness": 4.281409740447998, "identity": 0.6615633368492126, "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_NEWS_reg/4db15cd6f3d16f431c0d287d9092b4fd4b85b45586755c9019c40faa64e39c84.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": "4dd78d05bb9e75f8b2a5421184d8dda11756554bf68a6b6e673bc172a083c125", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Escribí el número en el sobre.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0003471382296580817391309654480dcce102a07885d7b63024ff902da59905", "generation_seconds_batch": 15.709006855031475, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1891307830810547, "target_z": 1.2734375363398547, "ce": 6.316492557525635, "genuineness": 4.344170093536377, "identity": 0.7707008123397827, "hypothesis": "Escribí el número en el sobre. Lélelo bien y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/4dd78d05bb9e75f8b2a5421184d8dda11756554bf68a6b6e673bc172a083c125.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "4ef2e84ecf61719a69f53ad69bf6496268583ca7d899ea00cef9d1fc30960b9d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d0d6d74db943be47d12a84275eefb6b39594efd1e155615820cbd49e58d7bfe1", "generation_seconds_batch": 16.22077202005312, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.417281150817871, "target_z": 1.6874999121082583, "ce": 6.268836498260498, "genuineness": 4.3504462242126465, "identity": 0.7417248487472534, "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_NEWS_reg/4ef2e84ecf61719a69f53ad69bf6496268583ca7d899ea00cef9d1fc30960b9d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "4f632bded00c11dca4be23ef1f008d3230d9c4c9b969caf1562e7f61660152ac", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8b43ae6dd837e7412260e9b4b1309f39f4d67a80d231feb300f6397903cc2ae0", "generation_seconds_batch": 18.803161473944783, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9510672092437744, "target_z": 2.656250152120322, "ce": 5.998785495758057, "genuineness": 3.9048471450805664, "identity": 0.4075867831707001, "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_NEWS_reg/4f632bded00c11dca4be23ef1f008d3230d9c4c9b969caf1562e7f61660152ac.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "4f7e6846a363d40504180acbaab6f2cee16555650185b75d13b01359aa909c47", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "eb164e9b94b4a3044abd496f0517dae057d78ca92049321741bfedb628eac842", "generation_seconds_batch": 18.607872123015113, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5509552359580994, "target_z": 0.11523434859022187, "ce": 6.41180419921875, "genuineness": 4.331617832183838, "identity": 0.5282473564147949, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/4f7e6846a363d40504180acbaab6f2cee16555650185b75d13b01359aa909c47.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "4f882a74223a1432e4098b4d7b9447b947bba172b0adc424b81f6e797505d91b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8c8daeac5fe20e50d421d2bfb90501fd871d48faaf9821bcfece8d560354dbd3", "generation_seconds_batch": 19.951832809019834, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5636417865753174, "target_z": 1.953124873233065, "ce": 5.974957466125488, "genuineness": 4.15588903427124, "identity": 0.7743780612945557, "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_NEWS_reg/4f882a74223a1432e4098b4d7b9447b947bba172b0adc424b81f6e797505d91b.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": "50b06ba49d9849e0ec40b2f787c66431ca8f8f5dfde9a127fec652c9451c5a99", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3327b675f9b6fb65a53b1c8c5fb2eddbb0ca141b9ea0ebe7673c46d9a505ceb2", "generation_seconds_batch": 13.974739265977405, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.226569652557373, "target_z": 3.1562502062075475, "ce": 6.316492557525635, "genuineness": 4.068024158477783, "identity": 0.6455157399177551, "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_NEWS_reg/50b06ba49d9849e0ec40b2f787c66431ca8f8f5dfde9a127fec652c9451c5a99.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "51813ea3c35763e4e46b688dd2474d519ef5d8e63b860c31f20aacc303e70d79", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ae8258139972b55c829a01ff15be25ece0ed5fe08e1f01aaf0683e4425e5f8a8", "generation_seconds_batch": 14.891127925016917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6927835941314697, "target_z": 2.187499966195484, "ce": 5.863759994506836, "genuineness": 3.9173994064331055, "identity": 0.6047097444534302, "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_NEWS_reg/51813ea3c35763e4e46b688dd2474d519ef5d8e63b860c31f20aacc303e70d79.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "51b233cf7caa0919be5717b5a4d26bfd3e80754f77dbff620efdc380b264f3fa", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9aec9ce9ded886e6cc7cf26db0ec1786ca6a7640e050e6a5226eb7c23f7251a9", "generation_seconds_batch": 14.665501038078219, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8555145263671875, "target_z": 0.6679686904195405, "ce": 6.221180438995361, "genuineness": 3.9048471450805664, "identity": 0.15310119092464447, "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_NEWS_reg/51b233cf7caa0919be5717b5a4d26bfd3e80754f77dbff620efdc380b264f3fa.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": "5281349283ee35a79be700a69c26febda95628c7567b032580c49209a3663238", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ab9a0ffd238928378612b5b9c3363a78e6a24d15ad8262df3317e3eeda9c4fb7", "generation_seconds_batch": 14.542624247027561, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.493462562561035, "target_z": 3.640625109864677, "ce": 6.41180419921875, "genuineness": 3.9299514293670654, "identity": 0.4041714072227478, "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_NEWS_reg/5281349283ee35a79be700a69c26febda95628c7567b032580c49209a3663238.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "52909829ef582acee06f3901cb2c2b2cbe0fc4178a655385b29918e717fb8096", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fe6983ac3563649ab103f04b51064fb9cf0739d6f9a5368388800ee2f1722d50", "generation_seconds_batch": 20.546833553002216, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4258906841278076, "target_z": 1.7031250625383545, "ce": 6.221180438995361, "genuineness": 4.243753433227539, "identity": 0.41282549500465393, "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_NEWS_reg/52909829ef582acee06f3901cb2c2b2cbe0fc4178a655385b29918e717fb8096.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "52b31b7819565233edb47b57b9762e955f892daf274af85b8aaa012c26042b24", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "37b11dcd1ff98721cebadc921cd9729f5017c3528c8e175864a84b458e2279b5", "generation_seconds_batch": 19.5266365269199, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.025551199913025, "target_z": 0.9765624907037581, "ce": 6.205295085906982, "genuineness": 3.8169827461242676, "identity": 0.21124263107776642, "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_NEWS_reg/52b31b7819565233edb47b57b9762e955f892daf274af85b8aaa012c26042b24.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "52fc37bff27186459b327c0c78d842eadd2c27909e41884febe9559e5e61f53e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7702cf9ed231b25b378a6597e066a77f898293f333175b5a7096581667c6fbbd", "generation_seconds_batch": 20.897332945023663, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2537016868591309, "target_z": 1.3906250828210642, "ce": 6.252951145172119, "genuineness": 3.7040140628814697, "identity": 0.3031056821346283, "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_NEWS_reg/52fc37bff27186459b327c0c78d842eadd2c27909e41884febe9559e5e61f53e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "53d99f4228b317cf5fe57356f07a10ccfcf1d358136e0489fee06493750ee8c6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fbb5aeba306b73bb3d6295d60ba2cafec4866e3d896295301553d8a2f2510661", "generation_seconds_batch": 16.69160560402088, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9911134243011475, "target_z": 0.9140625380300805, "ce": 6.316492557525635, "genuineness": 4.519899368286133, "identity": 0.6827431917190552, "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_NEWS_reg/53d99f4228b317cf5fe57356f07a10ccfcf1d358136e0489fee06493750ee8c6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "53ebed057ebc88f752b3654e6f55a33585491f06e53e2cc3625ff59cfcda947c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c5475495661abd87806783d570feade57069745aa731e603a4035b474a980a0a", "generation_seconds_batch": 16.22077202005312, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.572251319885254, "target_z": 1.9687500236631612, "ce": 6.1894097328186035, "genuineness": 4.406930446624756, "identity": 0.6961395740509033, "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_NEWS_reg/53ebed057ebc88f752b3654e6f55a33585491f06e53e2cc3625ff59cfcda947c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "54035e55bf9500f7a4614f67504e38f6bfb3797e9220fb873b9ec85051f5e333", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e9d557fecd8c4cc920cd11b4ea6ed887d185c557d5282bbc3f7b76effd36119a", "generation_seconds_batch": 14.757018235046417, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6755647659301758, "target_z": 2.1562500980330963, "ce": 6.38003396987915, "genuineness": 4.193545341491699, "identity": 0.5856651067733765, "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_NEWS_reg/54035e55bf9500f7a4614f67504e38f6bfb3797e9220fb873b9ec85051f5e333.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "542abe5827d6c9de32d6a7c1e8074a5708defa28b3c52f152077ed4d4db316b7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "93e6e0b9390f75b00a48ff9b7d703093ef92cc3f727a32548039e719ee89ecc7", "generation_seconds_batch": 18.972911934019066, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 6.09409761428833, "genuineness": 4.306513786315918, "identity": 0.7007701396942139, "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_NEWS_reg/542abe5827d6c9de32d6a7c1e8074a5708defa28b3c52f152077ed4d4db316b7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "543d105c61c05088060637eead09cb4d0f719285f37e74d95eed09bcfd2ab4d4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ed1ff512dc51d155159066f34df6175c771b15eaae6b6e016816509aa043cc2a", "generation_seconds_batch": 19.14715797593817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0018751621246338, "target_z": 0.9335936515443471, "ce": 6.324435234069824, "genuineness": 4.337893962860107, "identity": 0.29910120368003845, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/543d105c61c05088060637eead09cb4d0f719285f37e74d95eed09bcfd2ab4d4.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "2eda68f6d063d23207661e94cb541533f2694d560aa21151c940afe1a5de0c59", "generation_seconds_batch": 14.41017354291398, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9855048656463623, "target_z": 2.7187498884450974, "ce": 6.2767791748046875, "genuineness": 3.9927117824554443, "identity": 0.520461916923523, "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_NEWS_reg/546203f143d2dbd908be3097fd76d41d76e93afd1487fa15c64a3d119dff5d49.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "55170188f5d77b2b8d055166fb33a8c672921770a38b0c71c655e51934d255a5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "92c5db2451bee7dfc7991e019dad3b54e21edaa64b1ce28fe440def84a9aa498", "generation_seconds_batch": 20.95940620091278, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7100024223327637, "target_z": 2.2187498343578715, "ce": 6.332377910614014, "genuineness": 4.168440818786621, "identity": 0.45335686206817627, "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_NEWS_reg/55170188f5d77b2b8d055166fb33a8c672921770a38b0c71c655e51934d255a5.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": "561dbebd497f9d3b36fe980872e8c98c0efc55c048acee5bdf2766c39d166304", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "997ea0b39d051760e29fcfa143d6dd71a695c445295a53052856893951aadb64", "generation_seconds_batch": 19.668307394022122, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.123256206512451, "target_z": 2.9687501318376124, "ce": 6.006728172302246, "genuineness": 3.9048471450805664, "identity": 0.7454414367675781, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois. Je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/561dbebd497f9d3b36fe980872e8c98c0efc55c048acee5bdf2766c39d166304.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "57059530b691b7dfaea52a3ae146ced3dc47c7528ca2e7a17aba05780d457120", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "16117c4620d0cc6c086e1ac1d657c15c919c8c33c412fb796d7aaf98b8eaeb9b", "generation_seconds_batch": 20.976721537066624, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7737247943878174, "target_z": 0.5195312757759435, "ce": 6.149696350097656, "genuineness": 4.268857479095459, "identity": 0.8244615197181702, "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_NEWS_reg/57059530b691b7dfaea52a3ae146ced3dc47c7528ca2e7a17aba05780d457120.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": "571024783b1fc55d242f316e5170b1affaa7be1f227e10f78b870b9db790145c", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The road turns left behind the bridge.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6a5cb44af90fca00a3608cd046d8830a684f2728936af3a6911a16fc0897fa28", "generation_seconds_batch": 14.21999103098642, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.503375768661499, "target_z": 1.843750118315806, "ce": 6.356205940246582, "genuineness": 3.810706615447998, "identity": 0.13690121471881866, "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_NEWS_reg/571024783b1fc55d242f316e5170b1affaa7be1f227e10f78b870b9db790145c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "571d1e20a1c53490cc0346c2c7bbd8564bd5d2f0d60fa40cd7083c8c6fb4545e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e9c1dfa77ab9c48a55efbba5a3f3f94849438853adf1a2db32fe18435b5c6001", "generation_seconds_batch": 18.66631546197459, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3311867713928223, "target_z": 1.5312501385985156, "ce": 6.054384231567383, "genuineness": 4.306513786315918, "identity": 0.7576913833618164, "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_NEWS_reg/571d1e20a1c53490cc0346c2c7bbd8564bd5d2f0d60fa40cd7083c8c6fb4545e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "572f832e714e8a0fa7c3e85bcb92a5098c996dda20c5073b41b6ee48eddb0cf5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "248a1c24ad9d5edc246db999704e669b21e31a93be4a0fefa212763a77da9712", "generation_seconds_batch": 19.121506287017837, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8477537631988525, "target_z": 2.468750077750387, "ce": 6.284721851348877, "genuineness": 4.293961524963379, "identity": 0.6097174286842346, "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_NEWS_reg/572f832e714e8a0fa7c3e85bcb92a5098c996dda20c5073b41b6ee48eddb0cf5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "57734ceca8f5ee17d652231265227a4a2bf7d5dd6def7c5b908c6871e227d1f1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e33e744a5b13477ceb9a889d2a4ad30db8833d83826eba2da4638c745062b119", "generation_seconds_batch": 19.215223321923986, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.078212261199951, "genuineness": 4.243753433227539, "identity": 0.6457660794258118, "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_NEWS_reg/57734ceca8f5ee17d652231265227a4a2bf7d5dd6def7c5b908c6871e227d1f1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "5814f02772596ab4b1ef57ed875449151fb14a02eed0fa9eabba41bf938a686b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "71f54f79e9073d49caf4361f3f1edb4495a08ad5a1354d88eda34a6814701ea3", "generation_seconds_batch": 18.902449395973235, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0363130569458008, "target_z": 0.9960938205669272, "ce": 6.252951145172119, "genuineness": 4.469690799713135, "identity": 0.5724173188209534, "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_NEWS_reg/5814f02772596ab4b1ef57ed875449151fb14a02eed0fa9eabba41bf938a686b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "5820774252ef107ec66333ce8f4e195b0ac4dc566a1fd86fe4c732ab93bf61f7", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "93f5dec6376cfd19cc5ea759435df69b6ee11791d2a018c0b496aa51b545dc37", "generation_seconds_batch": 14.591332069016062, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9911134243011475, "target_z": 0.9140625380300805, "ce": 6.268836498260498, "genuineness": 4.419482707977295, "identity": 0.6337631940841675, "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_NEWS_reg/5820774252ef107ec66333ce8f4e195b0ac4dc566a1fd86fe4c732ab93bf61f7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "5895960a89fc0fa944964a7d71d2313ca067284ef9a2a1d97be5ed4cacf9a0ef", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d683257fff3b4de53a604614cd466823c5d57b8d8f51555e55da6c302b926a5a", "generation_seconds_batch": 20.423111359938048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.41180419921875, "genuineness": 4.080576419830322, "identity": 0.5376282930374146, "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_NEWS_reg/5895960a89fc0fa944964a7d71d2313ca067284ef9a2a1d97be5ed4cacf9a0ef.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": "589ca0d94af6a35113fbaced027cf3ee22caca0d9c7e1c3057c953a753a8831c", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "fde729a8354444f38c2cbfdb2bdd74f9bb1781f298880621d0c106ba3f5b4f15", "generation_seconds_batch": 14.289620777009986, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.347101926803589, "target_z": 3.37500014873987, "ce": 6.252951145172119, "genuineness": 4.143336772918701, "identity": 0.43758660554885864, "hypothesis": "Jemand hat den kleinen blauen Stuhl umgestellt. Er stand hier neben dem hohen Bücherregal. Wir stellen ihn nach dem Essen zurück.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/589ca0d94af6a35113fbaced027cf3ee22caca0d9c7e1c3057c953a753a8831c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "58a54b98223c0387b79f33755583d75b91f6caee75a1859bd4828b02e56493da", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5cd061a88ff98a413c363151ff3090578057bff395d1547d334bdb5cd6a1a8f5", "generation_seconds_batch": 20.653730508987792, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6669552326202393, "target_z": 2.1406249476030004, "ce": 6.395919322967529, "genuineness": 4.18099308013916, "identity": 0.5885293483734131, "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_NEWS_reg/58a54b98223c0387b79f33755583d75b91f6caee75a1859bd4828b02e56493da.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "58c41d8701fa8e36955504e47d46ad252c141e9c6345d90666580174013e05fe", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ed2ad18cedb19d91117011cd6a2b582f80d66d864c5abdd455cc030cbb00e674", "generation_seconds_batch": 19.248854364966974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6327450275421143, "target_z": 0.2636718714082702, "ce": 6.292664527893066, "genuineness": 4.137060642242432, "identity": 0.25517207384109497, "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_NEWS_reg/58c41d8701fa8e36955504e47d46ad252c141e9c6345d90666580174013e05fe.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "58e7bb495d2e61305c1efdd63910625c693ef3d89fadfa10d66170dc610e7a2f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e75d670d6fa8aaa853ca6c8a162230a71269e2d621a96b602d33358d6929e3d7", "generation_seconds_batch": 21.045267187058926, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5593371391296387, "target_z": 1.9453125143669192, "ce": 6.316492557525635, "genuineness": 4.218649387359619, "identity": 0.36714476346969604, "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_NEWS_reg/58e7bb495d2e61305c1efdd63910625c693ef3d89fadfa10d66170dc610e7a2f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "591a11aacc3e28fe2ef3e3e97de017e8bed72db8acad03a432c9003d61b746b1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a1a1b0710dd62109ad1e8c67110fd3d658b7a28eea2d4bcfa7e550f3ae1209fd", "generation_seconds_batch": 19.338424205081537, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.873582124710083, "target_z": 2.5156250963428706, "ce": 6.125868320465088, "genuineness": 4.193545341491699, "identity": 0.369062602519989, "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_NEWS_reg/591a11aacc3e28fe2ef3e3e97de017e8bed72db8acad03a432c9003d61b746b1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "592317897859fdf323d0b957fbc544301f0a75b8289ee975b53ff7230920fa61", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5513b10aded19ddeb42b823396c206543cd034ebe7bc2f6c4119bb07aaca7300", "generation_seconds_batch": 14.93867017608136, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6833255290985107, "target_z": 0.35546871070225017, "ce": 6.157639026641846, "genuineness": 4.13078498840332, "identity": 0.2723213732242584, "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_NEWS_reg/592317897859fdf323d0b957fbc544301f0a75b8289ee975b53ff7230920fa61.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "5938e6c49f33a56273b1e3c39fe29d9cd02ee504a50f04f3ea827594e7b99d7e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0045e172c73c88c9b68212dd5fabc94bbb98a011a05362a58d9ff6635833e404", "generation_seconds_batch": 14.485948589979671, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8133158683776855, "target_z": 2.4062499087278066, "ce": 6.41180419921875, "genuineness": 4.068024158477783, "identity": 0.47492459416389465, "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_NEWS_reg/5938e6c49f33a56273b1e3c39fe29d9cd02ee504a50f04f3ea827594e7b99d7e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "595c5bc53937cce59520ef20a4e325849ecd5a278c4a135fe230474a9dedcb4d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d6da3a7d1a77fb01ecfb305ba85e55812db994dddb1ade233dc76a11df4e06e8", "generation_seconds_batch": 20.057255571940914, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.451719045639038, "target_z": 1.7500000811308385, "ce": 6.173524379730225, "genuineness": 4.344170093536377, "identity": 0.731014609336853, "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_NEWS_reg/595c5bc53937cce59520ef20a4e325849ecd5a278c4a135fe230474a9dedcb4d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "59cc026fadd6a3d895bc028d521d13758d7b58219c5beb52d5816372c7b92c30", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "89a061cb4819307c41d9769c95e6c6b7604deaeb05e7d3744cd4bf2740664901", "generation_seconds_batch": 18.842281606979668, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6117594838142395, "target_z": 0.2255859188018771, "ce": 6.372091293334961, "genuineness": 4.507347106933594, "identity": 0.4243699908256531, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/59cc026fadd6a3d895bc028d521d13758d7b58219c5beb52d5816372c7b92c30.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "5a0383afa92eda3c60deee7615d4a541fd23531bb47f4cdef48bcaadaf4ab386", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7caae1c80adc7bb0e079405b82822d28c5efecd62720b434f6f551589d17e679", "generation_seconds_batch": 14.5518639809452, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.133169412612915, "target_z": 1.1718751402887415, "ce": 6.0305562019348145, "genuineness": 4.193545341491699, "identity": 0.6365584135055542, "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_NEWS_reg/5a0383afa92eda3c60deee7615d4a541fd23531bb47f4cdef48bcaadaf4ab386.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "5ad37a1e7e37fd0b13e35f99dacbe923e2ececb9ee86f8d64bca98875ba970a9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6103dfe1cd698b17a8d44fe35938b1bca4e2f66f2fbd02f0dde3194788149a20", "generation_seconds_batch": 20.33513457106892, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5292041301727295, "target_z": 1.8906251369082898, "ce": 6.133810997009277, "genuineness": 4.1998209953308105, "identity": 0.6919549703598022, "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_NEWS_reg/5ad37a1e7e37fd0b13e35f99dacbe923e2ececb9ee86f8d64bca98875ba970a9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "5b1e12044b0589762aa38e0f352628192c8780f8f53c86e3899b4e7ff1e1e413", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "20699a74aba0d15f98971a4810f2e8cc6f6fb20e52d7fb92163f4fce4315d711", "generation_seconds_batch": 20.597606427036226, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5636417865753174, "target_z": 1.953124873233065, "ce": 6.109982967376709, "genuineness": 4.331617832183838, "identity": 0.6844422817230225, "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_NEWS_reg/5b1e12044b0589762aa38e0f352628192c8780f8f53c86e3899b4e7ff1e1e413.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "5b40da5db8aafaa72670e72ab7451ef793e1ef4dde2d4ead57630c45989a3558", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ae4c943897987fc0a6f07fbfc0c0ef9f4111e4caad9d95cb83703178e5b721a9", "generation_seconds_batch": 19.121506287017837, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9609802961349487, "target_z": 0.8593749442225486, "ce": 6.292664527893066, "genuineness": 4.7081804275512695, "identity": 0.6937456727027893, "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_NEWS_reg/5b40da5db8aafaa72670e72ab7451ef793e1ef4dde2d4ead57630c45989a3558.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": "5c221945debd29ae887cd09cb2001f684245c229528a3b576ff7ef430c7e81cb", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Escribí el número en el sobre.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "03a3df1a1c4d4455c93b6a461130ba9b566f1b412ad80ce31a20bbf95e454a9b", "generation_seconds_batch": 14.610306482063606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.873582124710083, "target_z": 2.5156250963428706, "ce": 6.157639026641846, "genuineness": 4.13078498840332, "identity": 0.6754097938537598, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/5c221945debd29ae887cd09cb2001f684245c229528a3b576ff7ef430c7e81cb.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": "5c2e37419b9252e434ab72405c31b5e0ad6fbbde278d12444fe035a7a083b37d", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "1d902b274b9525301494f278c4a2d0a290474a806c39d235335b5aa5b1fc7434", "generation_seconds_batch": 14.118994012009352, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2795300483703613, "target_z": 1.437500101413548, "ce": 6.070269584655762, "genuineness": 3.923675537109375, "identity": 0.7899686694145203, "hypothesis": "La route tourne à gauche après le pont. Continue jusqu'à la petite barrière en bois. Je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/5c2e37419b9252e434ab72405c31b5e0ad6fbbde278d12444fe035a7a083b37d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "5cb283a8ba4fc14f026791a392ef10e22e176082a9643d6870b6d3f0bfd580bf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7d4768b6b383d245ab7fcab5224c69292847d51128eb62fdbc33e5f03f941045", "generation_seconds_batch": 20.867851946968585, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6239080429077148, "target_z": 2.0625000608481288, "ce": 5.990842819213867, "genuineness": 4.042920112609863, "identity": 0.6416593790054321, "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_NEWS_reg/5cb283a8ba4fc14f026791a392ef10e22e176082a9643d6870b6d3f0bfd580bf.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": "5cbdbb92edcf6c0e9cb13750e251235104443737588b89372a7ece7923333f6a", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "e0180eee1cd10f65114ce4ac8000ddea58c21eb9b143874e903acaf4f8d9c8b1", "generation_seconds_batch": 14.929697441053577, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.149696350097656, "genuineness": 4.061748504638672, "identity": 0.5743850469589233, "hypothesis": "Tu mensaje llegó después de que me fuera. Lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/5cbdbb92edcf6c0e9cb13750e251235104443737588b89372a7ece7923333f6a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "5d4d4f547fbaf5d0c442eec17b7783bfdff24eba6fc4f767a56c1970913f82af", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "c272cec6b65e0e51a39ac3984bbccce849965158a66e26a73a8648e84e43e0c3", "generation_seconds_batch": 14.389843570068479, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9997228384017944, "target_z": 0.9296874721112743, "ce": 6.515059471130371, "genuineness": 4.20609712600708, "identity": 0.4865836501121521, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/5d4d4f547fbaf5d0c442eec17b7783bfdff24eba6fc4f767a56c1970913f82af.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "5d52b6b704c2a013417418aba0eb72351f7f297878735dc6bda2f07ca635bdc8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2be2119055a93d8dd0932a1f2b4e902652387bdd67253967ad46ad3165c970e6", "generation_seconds_batch": 14.399691935046576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7091538906097412, "target_z": 0.40234372929473394, "ce": 6.292664527893066, "genuineness": 4.143336772918701, "identity": 0.4005274474620819, "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_NEWS_reg/5d52b6b704c2a013417418aba0eb72351f7f297878735dc6bda2f07ca635bdc8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "5d6d4c40e25fdcd18223ce66f637105c010b5a91281f017bf022ce614af8c680", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e46a4d496e675f9280959fb0770d0f6fb59803ac1ef34b973d6fbbb32a4eba20", "generation_seconds_batch": 18.71039308293257, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2278733253479004, "target_z": 1.3437500642285805, "ce": 6.062326908111572, "genuineness": 4.15588903427124, "identity": 0.48612600564956665, "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_NEWS_reg/5d6d4c40e25fdcd18223ce66f637105c010b5a91281f017bf022ce614af8c680.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "5d7a86abac186725082a74d8d8e7f9f1dab4afb114feb009e54e11806699b052", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "09bfcabbfc972b4b5caf98b5596dc96817a09961245bfff3353d1ee6129ecc3f", "generation_seconds_batch": 19.005299902986735, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0341606140136719, "target_z": 0.9921874247849519, "ce": 6.324435234069824, "genuineness": 4.482243061065674, "identity": 0.7615740299224854, "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_NEWS_reg/5d7a86abac186725082a74d8d8e7f9f1dab4afb114feb009e54e11806699b052.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "5d8e04f74845d672a2ac4104890d694c242ad65e14990fea7fdbb72cfe1250bf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e3aef6beff4c0fa6a3c18f39c3f8142aac912cff6c87f8ece9b1d5fffb973b25", "generation_seconds_batch": 18.96763893100433, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0642937421798706, "target_z": 1.0468750185924838, "ce": 6.24500846862793, "genuineness": 4.595211505889893, "identity": 0.6295630931854248, "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_NEWS_reg/5d8e04f74845d672a2ac4104890d694c242ad65e14990fea7fdbb72cfe1250bf.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "2529386e9cda2a364aca200ffdb3bb9ccd85510f3583edaf107af4c6f79dcf43", "generation_seconds_batch": 13.889720431994647, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0772079229354858, "target_z": 1.0703125278887258, "ce": 6.046441555023193, "genuineness": 4.193545341491699, "identity": 0.7143406867980957, "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_NEWS_reg/5e0af124ad7421917e897f2876d49b25e213a1bc87e01fe43f09b9c255c2a32f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "5e67f71dee40b7027b3d42ebc65d92f1cf446d8725f493256fbf87cb36ac1a09", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e5617a5f5262cd0626fa8b37fc7e59ce5fc07c73518625a5a2708b677d910697", "generation_seconds_batch": 20.65841965796426, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3096630573272705, "target_z": 1.4921874788721774, "ce": 6.197352409362793, "genuineness": 4.193545341491699, "identity": 0.49138179421424866, "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_NEWS_reg/5e67f71dee40b7027b3d42ebc65d92f1cf446d8725f493256fbf87cb36ac1a09.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "5e81f76c8be059ff2d09931ac7a88b3be52a88b7507e8a60b532ab34fcebb530", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fa61562b08ab4b484d6d270f1cc90c047ea4275e1bdcc128cba28b3549bff866", "generation_seconds_batch": 18.800731925992295, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5464229583740234, "target_z": 1.9218750050706774, "ce": 6.086154937744141, "genuineness": 4.419482707977295, "identity": 0.7670502066612244, "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_NEWS_reg/5e81f76c8be059ff2d09931ac7a88b3be52a88b7507e8a60b532ab34fcebb530.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "5ea321ddb8c07ab71d5967834285340056003afd1b288ab0ffd66a6af8b5a35c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d4fe77432bfcf36af5f8f2601e753168cc152fec029beb04059160bb58734f3f", "generation_seconds_batch": 20.006567763979547, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5119850635528564, "target_z": 1.8593748360480973, "ce": 6.2767791748046875, "genuineness": 4.193545341491699, "identity": 0.5243796110153198, "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_NEWS_reg/5ea321ddb8c07ab71d5967834285340056003afd1b288ab0ffd66a6af8b5a35c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "5ed275ee7e82ff9e65db3e69dabb07b750cb5a2e72dfd10cd5f15561eddfa245", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "aa0f9f072c39da61495012d3c1f5769ef7a8d65540e643ee52044ba8aa1e8676", "generation_seconds_batch": 14.26074289297685, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6897826194763184, "target_z": 0.3671874653503711, "ce": 6.41180419921875, "genuineness": 4.375550270080566, "identity": 0.46172958612442017, "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_NEWS_reg/5ed275ee7e82ff9e65db3e69dabb07b750cb5a2e72dfd10cd5f15561eddfa245.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "5f087c77402f9f3d5bab5f95db492058ef8f5369c5f343d64bdc121d177244f8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ad2c20cd84dd3aaf59ab9cc0aae583a8e1c4cd22248e07b0886019bbc2e872ca", "generation_seconds_batch": 20.16473301802762, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2450921535491943, "target_z": 1.374999932390968, "ce": 6.268836498260498, "genuineness": 4.457139015197754, "identity": 0.7248646020889282, "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_NEWS_reg/5f087c77402f9f3d5bab5f95db492058ef8f5369c5f343d64bdc121d177244f8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "5f66c1e126c999e4a59d428480b21c7f1d15def9f50562941f3a92829760bec1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2cb3e3811709aa3049f496b912c57cecdc224c615b95cd94fac6edc0cf4b1724", "generation_seconds_batch": 21.259091073996387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5724788904190063, "target_z": 0.15429690014210878, "ce": 6.252951145172119, "genuineness": 4.3127899169921875, "identity": 0.2864159047603607, "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_NEWS_reg/5f66c1e126c999e4a59d428480b21c7f1d15def9f50562941f3a92829760bec1.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": "5f80ea819254663e05aa33c2f022868d707cce13be7f1d99727d4ae35d19bddb", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "49f5158bb64d3f34fe32af156f1beb0d20377c7c2e5c854450ed6f80ac0f9fbc", "generation_seconds_batch": 14.704422302078456, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.192131757736206, "target_z": 3.0937500371849675, "ce": 6.062326908111572, "genuineness": 4.024092197418213, "identity": 0.6384303569793701, "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_NEWS_reg/5f80ea819254663e05aa33c2f022868d707cce13be7f1d99727d4ae35d19bddb.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "5fd6d5a738e6832762c1d816c4f9843e8f3758534f0e4e9889baeeb8d8baf3cd", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "928b7cd141fc834b9b1515b99b8ed9f78e1587b3f3d06999831d4a08213ff005", "generation_seconds_batch": 15.473690992104821, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8017054796218872, "target_z": 0.5703124738015001, "ce": 6.419747352600098, "genuineness": 4.331617832183838, "identity": 0.4726965129375458, "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_NEWS_reg/5fd6d5a738e6832762c1d816c4f9843e8f3758534f0e4e9889baeeb8d8baf3cd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "5fd71e7bfaf4d0d9fb14113b473ff6911c4df166ae8c750e27e530053f715808", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0784bb1cc1802cf2c74361cff3c7b152f57d72c7f8509648748b3f089faffa3a", "generation_seconds_batch": 15.157238428015262, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7186119556427002, "target_z": 2.234374984787968, "ce": 6.459460258483887, "genuineness": 4.193545341491699, "identity": 0.415427029132843, "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_NEWS_reg/5fd71e7bfaf4d0d9fb14113b473ff6911c4df166ae8c750e27e530053f715808.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "60114f4c01eb417faeae5f637630ae9161dbe1c86b8191f85ec3bd863a2e4783", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b0300c2b58ff6b6e23e64e0afa26c6b3fb69c0c09fdca86836a65b33ffa5504e", "generation_seconds_batch": 21.00167818600312, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9265425205230713, "target_z": 0.796874991548871, "ce": 6.403861999511719, "genuineness": 4.3127899169921875, "identity": 0.32645320892333984, "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_NEWS_reg/60114f4c01eb417faeae5f637630ae9161dbe1c86b8191f85ec3bd863a2e4783.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": "6014ee3992a94f49d338dce32ad576bc15e0166d36b0bfd14d7c058a8e2b3c82", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Your message arrived after I had left.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4e08ccbaf8d0f4e70eb436b318f757fa78356d3271b6a78e9296e479d9709ee2", "generation_seconds_batch": 20.496060185017996, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1848260164260864, "target_z": 1.2656249611248065, "ce": 6.292664527893066, "genuineness": 3.8420867919921875, "identity": 0.3902676999568939, "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_NEWS_reg/6014ee3992a94f49d338dce32ad576bc15e0166d36b0bfd14d7c058a8e2b3c82.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "609a39ad6fdb08eda6678a9e42a3c06d14e3929d1668aa93b1b212d0109ab68e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3f007feb3d86d6cdddabf4472a57aafb060fcbbcae807a15533a1da1214e1b00", "generation_seconds_batch": 20.32046227401588, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0457711219787598, "target_z": 2.828125076060161, "ce": 6.38003396987915, "genuineness": 3.8797430992126465, "identity": 0.25562888383865356, "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_NEWS_reg/609a39ad6fdb08eda6678a9e42a3c06d14e3929d1668aa93b1b212d0109ab68e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "60a9aafff5b8664b49226e0b5edd53ea2c5c3c01b53ff8df02f77054d4302615", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8c64c0c5166c74a8e5fb37727490fb4955da0065b60844cd3f524bcd6c30b7db", "generation_seconds_batch": 18.77553105202969, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.046441555023193, "genuineness": 4.557555198669434, "identity": 0.6850699186325073, "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_NEWS_reg/60a9aafff5b8664b49226e0b5edd53ea2c5c3c01b53ff8df02f77054d4302615.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "60b00a2ec8a1a670dc97e98ee171a29135252eab5214907220118833ae80670b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "99e2616944a63ce3572b57733e6d29e159cdc272afb463f08de091f06c91a5c7", "generation_seconds_batch": 15.839115922921337, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9695897698402405, "target_z": 0.8749999864781935, "ce": 6.149696350097656, "genuineness": 4.494794845581055, "identity": 0.7142200469970703, "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_NEWS_reg/60b00a2ec8a1a670dc97e98ee171a29135252eab5214907220118833ae80670b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "611626ec32d5c744e34dee61c9b2078cc24f14a23d476ee6961b49f2b4c23ea3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f815a71e1d646dfe78bf32db2f984046e5f2489fc0ba4f1a22f0cd4194af2423", "generation_seconds_batch": 14.941111410036683, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4775474071502686, "target_z": 1.7968750997233223, "ce": 6.340320587158203, "genuineness": 4.356721878051758, "identity": 0.5111196041107178, "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_NEWS_reg/611626ec32d5c744e34dee61c9b2078cc24f14a23d476ee6961b49f2b4c23ea3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "612f8baf8b57f0354f14a173a930608236ca5bdbdbaa18d8e9d0f7d76d0b53cd", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fb8e736e5f151a75c2ab4e382f74cc7fe7017596124b6177274106cb1553853d", "generation_seconds_batch": 18.959711924078874, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4000623226165771, "target_z": 1.6562500439458707, "ce": 6.125868320465088, "genuineness": 4.3881025314331055, "identity": 0.7092278003692627, "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_NEWS_reg/612f8baf8b57f0354f14a173a930608236ca5bdbdbaa18d8e9d0f7d76d0b53cd.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": "61810b542976c744e9c2f026e07d2deacbb5d9f791e6633f379a0b76b31fa9db", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "581864fa948d1e2bbe7df7500c55eda61e985f1285892c479540151e96707fc0", "generation_seconds_batch": 15.15650177502539, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3815395832061768, "target_z": 3.4374998850646454, "ce": 6.292664527893066, "genuineness": 3.961331605911255, "identity": 0.5910622477531433, "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_NEWS_reg/61810b542976c744e9c2f026e07d2deacbb5d9f791e6633f379a0b76b31fa9db.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": "61c45e2daa5edb587191305dafb0f43398d8d9a5da088e3e4cb81beed6beac25", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "cee4248e7ba7fd1b0f024a14d79f2e5b6f3907e16b29f7e5df0e5efc36e79188", "generation_seconds_batch": 14.84269108902663, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1934354305267334, "target_z": 1.2812498952060003, "ce": 6.1894097328186035, "genuineness": 4.20609712600708, "identity": 0.7090157270431519, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/61c45e2daa5edb587191305dafb0f43398d8d9a5da088e3e4cb81beed6beac25.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "61c57a0f4c3d8eebe8fe77250a6d247b25f9017a4a97fe0a1fe9f5253c8fbda2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0e5630392e804f8909110f0b9c23bf0d047650449fb9c6f712d9918a557b4679", "generation_seconds_batch": 19.25732411595527, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9596765041351318, "target_z": 2.6718748698526134, "ce": 6.3641486167907715, "genuineness": 4.15588903427124, "identity": 0.4091092050075531, "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_NEWS_reg/61c57a0f4c3d8eebe8fe77250a6d247b25f9017a4a97fe0a1fe9f5253c8fbda2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "6247f598c2e3b0d44d855c72b15ad73bf66b97fd337e2b302a450dcc46910a6f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "97e2fa351dda0526ca3e2009b8bd3b1af1f056fd91b01fdcec3886b185a81ef8", "generation_seconds_batch": 15.267664062906988, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0772079229354858, "target_z": 1.0703125278887258, "ce": 6.165581703186035, "genuineness": 4.406930446624756, "identity": 0.40934276580810547, "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_NEWS_reg/6247f598c2e3b0d44d855c72b15ad73bf66b97fd337e2b302a450dcc46910a6f.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": "628525cecf2c9bd7358ff02050374a4e99a0d0be40e7673efea156b19fcb7aa1", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "39d19f34f577ef9bb159fe5c4275eb91146cdbc280085224770e76a780786447", "generation_seconds_batch": 14.243553403066471, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1417787075042725, "target_z": 1.1874998580210328, "ce": 6.006728172302246, "genuineness": 4.099404335021973, "identity": 0.7958582639694214, "hypothesis": "La route tourne à gauche après le pont, continue jusqu'à la petite barrière en bois. Je t'attendrai de l'autre côté.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/628525cecf2c9bd7358ff02050374a4e99a0d0be40e7673efea156b19fcb7aa1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "62c76a73e7feeed79c94316cb1e3705d913522ce8430c02db7c968062cf8c356", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b0498cf9d49244647cb34febf565b25966f15613e9a156e49c94948ff3122c37", "generation_seconds_batch": 20.653730508987792, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.468937873840332, "target_z": 1.7812499492932259, "ce": 6.3641486167907715, "genuineness": 4.15588903427124, "identity": 0.2973766326904297, "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_NEWS_reg/62c76a73e7feeed79c94316cb1e3705d913522ce8430c02db7c968062cf8c356.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "63f7b74ac821955031c01223619fa256b4ba526cbe582faa5cd338f27834bbf2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ff7f555ce427f63efbedfe9e42d0b72989ffad2081ca6903cedb70a1d45a9433", "generation_seconds_batch": 20.593537491047755, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.086154937744141, "genuineness": 4.231201171875, "identity": 0.5615540742874146, "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_NEWS_reg/63f7b74ac821955031c01223619fa256b4ba526cbe582faa5cd338f27834bbf2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "6488066d6bf62f45a08883af11d4fc581c613577b55967935c6727fed07e4054", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f0dfc864f05f2b8957f113130ae8c18c56b1a828b3106a477f543ecbc158af4f", "generation_seconds_batch": 19.74287048506085, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.435632705688477, "genuineness": 4.18099308013916, "identity": 0.5809793472290039, "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_NEWS_reg/6488066d6bf62f45a08883af11d4fc581c613577b55967935c6727fed07e4054.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "64b07db570da327c1343d7353657c6c3a0398313547f21ea052098a5abbe5bc7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "40b399b8493fd869a6a876b3da3219c1b0f7e68e5838be4e61ae163c8033759f", "generation_seconds_batch": 20.1207884530304, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1245598793029785, "target_z": 1.1562499898586451, "ce": 6.284721851348877, "genuineness": 4.18726921081543, "identity": 0.2687572240829468, "hypothesis": "I found little box. It was beside the old window. We can look inside together.", "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_NEWS_reg/64b07db570da327c1343d7353657c6c3a0398313547f21ea052098a5abbe5bc7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "64da7f613c48ae92b94c0ea9d5d4c84d4183bb93945d9f5e200125ed7920c780", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "808f9b9f9b957fd65cb245a507b0d5b954101068ed1d593eadc3b3ff2584213c", "generation_seconds_batch": 18.707597031025216, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6497364044189453, "target_z": 2.1093750794406128, "ce": 6.157639026641846, "genuineness": 4.331617832183838, "identity": 0.6315382122993469, "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_NEWS_reg/64da7f613c48ae92b94c0ea9d5d4c84d4183bb93945d9f5e200125ed7920c780.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "650aad3ea2e74603d41e91db0980cf9de7836ff57ab85fcdfda1024807c7ed8f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3124b2d8a50aa50e22f7e20599ea78432d3a3441a31d7abd455f115c9ef05dc8", "generation_seconds_batch": 20.466894799028523, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8305349349975586, "target_z": 2.4375002095879994, "ce": 6.395919322967529, "genuineness": 3.9425034523010254, "identity": 0.2572859227657318, "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_NEWS_reg/650aad3ea2e74603d41e91db0980cf9de7836ff57ab85fcdfda1024807c7ed8f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "65939d6037bb067f83dbfca260a6508212996447a253a3ace214f3ae6dd24b01", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "814fbaafd0ff6fe6a71f6709cdab351dd211c0990fb668ee502b6a65a1555624", "generation_seconds_batch": 15.002767071011476, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.606689214706421, "target_z": 2.031250192685741, "ce": 6.340320587158203, "genuineness": 4.256305694580078, "identity": 0.5704309344291687, "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_NEWS_reg/65939d6037bb067f83dbfca260a6508212996447a253a3ace214f3ae6dd24b01.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "65a1529a111491b31aec8c4b1e806b39b1f08e5fbe210d03fc598273c609a49f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cef8c4c28fdef9399ad4dd38b8f0e700d5a58b304af2a8e95e2b7f196afcfd0b", "generation_seconds_batch": 19.28012532601133, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6755647659301758, "target_z": 2.1562500980330963, "ce": 6.403861999511719, "genuineness": 4.093128204345703, "identity": 0.4652717113494873, "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_NEWS_reg/65a1529a111491b31aec8c4b1e806b39b1f08e5fbe210d03fc598273c609a49f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "65a568143759e984531752b3a508fc1a103a63c9fb7b2ee61ba4fcb7f2d8130e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0264fd0ef8582861583d77c7b86a2c5cc9696d82718e6db775ca09e6a009ab7d", "generation_seconds_batch": 20.296078067971393, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3699291944503784, "target_z": 1.601562450138339, "ce": 6.173524379730225, "genuineness": 4.369274139404297, "identity": 0.7247891426086426, "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_NEWS_reg/65a568143759e984531752b3a508fc1a103a63c9fb7b2ee61ba4fcb7f2d8130e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "65f42f0b346918438eeca5f162e99f9663a3066a68fd35db32d6bb4591e444ea", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "899213398063041b341bdab6f6d3f45592c928f5e8f67cf8fa509ca4c9f9b2af", "generation_seconds_batch": 18.764511754969135, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2881394624710083, "target_z": 1.4531250354947418, "ce": 6.221180438995361, "genuineness": 3.8420867919921875, "identity": 0.25011226534843445, "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_NEWS_reg/65f42f0b346918438eeca5f162e99f9663a3066a68fd35db32d6bb4591e444ea.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "664eaa10b96c800f0e2b2f7d059a52ff0814d55cf668b9869124a52d39f4f871", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "72a6ac9d9a8ef482bc1952c681ca4e905111c92860757048224a0e7b07861e76", "generation_seconds_batch": 15.079204719979316, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.606689214706421, "target_z": 2.031250192685741, "ce": 6.260893821716309, "genuineness": 4.036643981933594, "identity": 0.7683899402618408, "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_NEWS_reg/664eaa10b96c800f0e2b2f7d059a52ff0814d55cf668b9869124a52d39f4f871.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "66cec3e048ad2f9e73e0986b197c6c2144823f7707cb7199f2af19f123e0fc2b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "dbe77804175e227d0919b1cc053ba6d9e323f7171d06eca4180193108eb9066a", "generation_seconds_batch": 18.82938202598598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3096630573272705, "target_z": 1.4921874788721774, "ce": 6.181467056274414, "genuineness": 4.494794845581055, "identity": 0.6965926289558411, "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_NEWS_reg/66cec3e048ad2f9e73e0986b197c6c2144823f7707cb7199f2af19f123e0fc2b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "673e974558d94471df728a0a784e709e90b1b05ecf40d9773655068013e7e518", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7e5a97015bbf69ee68da5cbccf3c5ee4d6a0c5ccd6096079f3614505f75b9c46", "generation_seconds_batch": 15.100355893955566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.232177972793579, "target_z": 1.351562423094726, "ce": 6.260893821716309, "genuineness": 4.3127899169921875, "identity": 0.6101839542388916, "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_NEWS_reg/673e974558d94471df728a0a784e709e90b1b05ecf40d9773655068013e7e518.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "67b73f76c2fda06fcc97f120a7cfdcb65900658f9f2c6322516d51752b36429b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.5, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0f3272db748d3e90068c090a07cffa69e66f3367a3f0c04a7d7b3d3c563d2de4", "generation_seconds_batch": 20.37464256596286, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.787487506866455, "target_z": 2.359374890135323, "ce": 6.356205940246582, "genuineness": 4.086852550506592, "identity": 0.49368202686309814, "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_NEWS_reg/67b73f76c2fda06fcc97f120a7cfdcb65900658f9f2c6322516d51752b36429b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "689f3d449015b19604fb263324d8cf8f6f0d537cec95ed8ae7dcb7de9e856619", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "934e2c3d568f97f63195922d62078117ef038a2d35c27621e90c148c030b914b", "generation_seconds_batch": 21.557112255948596, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0199427604675293, "target_z": 2.7812500574676773, "ce": 6.483288764953613, "genuineness": 4.042920112609863, "identity": 0.5736892223358154, "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_NEWS_reg/689f3d449015b19604fb263324d8cf8f6f0d537cec95ed8ae7dcb7de9e856619.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "68f354527960f592b2bb5da31f39d1e5b8b595a52fc1b71e3f777f71e321f265", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "5324e72ccaf1dfe81ea9c91bae28dd84330046b135902130039d8ce1dd6bd339", "generation_seconds_batch": 21.393760774983093, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9674373865127563, "target_z": 0.8710936988706696, "ce": 6.41180419921875, "genuineness": 4.193545341491699, "identity": 0.5437973737716675, "hypothesis": "I found the little box. It was beside the old window. We can look inside. Together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/68f354527960f592b2bb5da31f39d1e5b8b595a52fc1b71e3f777f71e321f265.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": "6960ac2f05b9aa1a62f42f61cf7eca81074fe957903eec5bc6a38ea2a726722e", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "d19f5780afb4ebc857ef0782f183dbcb3281dc1ea4f2329e99e9828af83db24a", "generation_seconds_batch": 19.39849733293522, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 6.24500846862793, "genuineness": 4.375550270080566, "identity": 0.7215709686279297, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/6960ac2f05b9aa1a62f42f61cf7eca81074fe957903eec5bc6a38ea2a726722e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "69ec9a93027fce5ec1ab20183981ce34b9c1acded9150cf7f013a5eb6964ae41", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4cefc931b46297654c3d71c26f622fdb1a9d0bfc4247d53afb74bdb85d5d6c34", "generation_seconds_batch": 19.123552117031068, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.054380416870117, "target_z": 2.8437497937924525, "ce": 5.967014789581299, "genuineness": 3.7291181087493896, "identity": 0.7777143716812134, "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_NEWS_reg/69ec9a93027fce5ec1ab20183981ce34b9c1acded9150cf7f013a5eb6964ae41.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "6a2c10ae79c38699cbaea3da2f133e97ff13db8051e6450c767be7f5a399fc27", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.75, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "04d26e200b3ee74b35803b3e34867322a8a65983ae193192406f2d71013659b7", "generation_seconds_batch": 19.098588880035095, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 5.998785495758057, "genuineness": 3.9801597595214844, "identity": 0.6385979652404785, "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_NEWS_reg/6a2c10ae79c38699cbaea3da2f133e97ff13db8051e6450c767be7f5a399fc27.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "6a4c727db7b1d107581a7e4523ba8e396dc58846b90b01ff49be29de611b0c5d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8810ec57000eb656858fc344f0bd46a92926f10b90125ea2e2b89c6e1d484074", "generation_seconds_batch": 19.53261306800414, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 6.229123115539551, "genuineness": 4.250029563903809, "identity": 0.4269951581954956, "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_NEWS_reg/6a4c727db7b1d107581a7e4523ba8e396dc58846b90b01ff49be29de611b0c5d.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "10bd6a2ea17111a5a667bb85f8182dadacaeb05a89326af9dc68904c9c4dd69d", "generation_seconds_batch": 14.802104649948888, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8318386077880859, "target_z": 0.625000067609032, "ce": 6.324435234069824, "genuineness": 4.093128204345703, "identity": 0.443103164434433, "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_NEWS_reg/6a61bc20d43b3e97640d8dd45682f7055c5dbaa5aa987cdb55f8a7d7a432a2c2.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": "6a6636db4ae8f95515ddb8c43d8304093491cac59a1c41b2f832282baa2d3521", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The road turns left behind the bridge.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] I will meet you on the other side.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The road turns left behind the bridge.\n(Natural clear speech.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Natural clear speech.) [3.2 seconds duration] I will meet you on the other side.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "81ec94e26787eea404829aa831753f6abc3699eaa48b4c33e4ecfaf8bdba44a8", "generation_seconds_batch": 15.90978264901787, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4129765033721924, "target_z": 1.6796875532421127, "ce": 6.340320587158203, "genuineness": 3.8797430992126465, "identity": 0.2549211382865906, "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_NEWS_reg/6a6636db4ae8f95515ddb8c43d8304093491cac59a1c41b2f832282baa2d3521.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "6a869373bb73422f476d470b9f8b989d38d0c09b3f6dc6fd91ac0af0a8f2fa89", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b971a228e86e685ff4476e858ce09ab4b9f48f7183010717885c77d2d3b6e758", "generation_seconds_batch": 20.360407141968608, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9459137916564941, "target_z": 0.8320312554932339, "ce": 6.38797664642334, "genuineness": 4.168440818786621, "identity": 0.20594358444213867, "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_NEWS_reg/6a869373bb73422f476d470b9f8b989d38d0c09b3f6dc6fd91ac0af0a8f2fa89.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "6b3481ee55f16fe0cfed80cb6a11535c5b132d6a01498d8a7dcead448415b7db", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "35d5dedf33fc7a9a7653d8f422e9b05d5a00f79e11f392a1c3d433d436b6eaa5", "generation_seconds_batch": 18.701014851918444, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.417281150817871, "target_z": 1.6874999121082583, "ce": 6.292664527893066, "genuineness": 4.469690799713135, "identity": 0.4581497013568878, "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_NEWS_reg/6b3481ee55f16fe0cfed80cb6a11535c5b132d6a01498d8a7dcead448415b7db.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": "6b4ccb1165594cdf9ae533ba6812d646443b5f62a63d6ee7d029ed203b6af9a1", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "c04cecdead180f12c922dd2ed977f5aac17779cd4e575181bd0e3df4c781ead5", "generation_seconds_batch": 20.16779401805252, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3441009521484375, "target_z": 1.5546876478947576, "ce": 6.046441555023193, "genuineness": 4.093128204345703, "identity": 0.699495255947113, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/6b4ccb1165594cdf9ae533ba6812d646443b5f62a63d6ee7d029ed203b6af9a1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "6b898acd9424ff43464a9d7736c0fefcca87947b6abe1c97c0e2b4a91c97bd13", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "db33cc0946dd228e01b2e5840089c7615001a779fa4f57a9d4a606f027cbea22", "generation_seconds_batch": 16.39072722697165, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.025551199913025, "target_z": 0.9765624907037581, "ce": 6.157639026641846, "genuineness": 4.519899368286133, "identity": 0.5788753628730774, "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_NEWS_reg/6b898acd9424ff43464a9d7736c0fefcca87947b6abe1c97c0e2b4a91c97bd13.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "d0d273dc2897ec5fb95dc7d63e08cc84cac929e51310ac2bfeb72016a71f66b6", "generation_seconds_batch": 15.394858321989886, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5079079866409302, "target_z": 0.03710935366089927, "ce": 6.23706579208374, "genuineness": 4.6705241203308105, "identity": 0.7109785079956055, "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_NEWS_reg/6b973c750e3616c70f28663c29496a0c74bea4a87dff4bad65cadc5394b99131.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "6bf4ea00fecde47162e587fec2e55a53b0e5ccb95e21e43f3c82a2ed821f6d0a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ac0e14fe0040e0192f02c343eb37feae1508f3204efdb4dcb7bcbdbfe14c6f36", "generation_seconds_batch": 21.038800382986665, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.047074794769287, "target_z": 1.0156249340811938, "ce": 6.2767791748046875, "genuineness": 4.017816066741943, "identity": 0.38730356097221375, "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_NEWS_reg/6bf4ea00fecde47162e587fec2e55a53b0e5ccb95e21e43f3c82a2ed821f6d0a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "6c39bf2aef5cb2f79e2a28e581c787e0602a9ff142c322690c3422e0168c0c83", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "09cb069f52700f7ad8262e5511adb886155ec1fe7fff859d55ac57744225910e", "generation_seconds_batch": 18.701014851918444, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4000623226165771, "target_z": 1.6562500439458707, "ce": 6.157639026641846, "genuineness": 4.306513786315918, "identity": 0.5787297487258911, "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_NEWS_reg/6c39bf2aef5cb2f79e2a28e581c787e0602a9ff142c322690c3422e0168c0c83.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "6c6a4e1475ee6ff59d979b1ea61c68553f216995002cc1770d3793580177d67f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3a0b03b1583893f248bc7fc86ad6177ddea7f36b79a05b2faa18921ef22b6ff7", "generation_seconds_batch": 18.867784505011514, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.606689214706421, "target_z": 2.031250192685741, "ce": 6.0146708488464355, "genuineness": 4.243753433227539, "identity": 0.6648436784744263, "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_NEWS_reg/6c6a4e1475ee6ff59d979b1ea61c68553f216995002cc1770d3793580177d67f.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": "6ccb4e5c63dddcc16bcc39d0ab0c37f197a425d350f25151594e122d3cd60e3a", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1c8aa369f5c0893a182792380a0a9407a6978461f6e33f1d068e963c54d78228", "generation_seconds_batch": 15.311326423077844, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.839144229888916, "target_z": 2.4531249273202906, "ce": 6.221180438995361, "genuineness": 4.3127899169921875, "identity": 0.6201783418655396, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de lau 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_NEWS_reg/6ccb4e5c63dddcc16bcc39d0ab0c37f197a425d350f25151594e122d3cd60e3a.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": "6ccc295ed47c34c91cc331c7ee1392261ca0d6191faa99eb203c40481912ccf4", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d98b882c5d3e314e6a11d977bd221951c9ba3b81439f9d3a8be979d7996706fd", "generation_seconds_batch": 19.701628713984974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0457711219787598, "target_z": 2.828125076060161, "ce": 6.046441555023193, "genuineness": 3.986435890197754, "identity": 0.5777629613876343, "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_NEWS_reg/6ccc295ed47c34c91cc331c7ee1392261ca0d6191faa99eb203c40481912ccf4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "6d6809c4f72cf1aecad2b57e0ceacd419e8bb8ebc81b85f630a88bb066378c25", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f2771c2aabe9d53a6e0921a6f34178460af080d8dc7a44fcb2eca7c35f10613b", "generation_seconds_batch": 20.73820419306867, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0715994834899902, "target_z": 2.875000094652645, "ce": 6.348263263702393, "genuineness": 3.8044304847717285, "identity": 0.5113006830215454, "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_NEWS_reg/6d6809c4f72cf1aecad2b57e0ceacd419e8bb8ebc81b85f630a88bb066378c25.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "6d6a6f12d496999666999a1315b60b3089f02292e8f04cf7a64ab700181da8bb", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3cd962704ba02f83480864de16deeff6a962c8b2f6651d37ebe73677c61a8175", "generation_seconds_batch": 21.40725150203798, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1245598793029785, "target_z": 1.1562499898586451, "ce": 6.157639026641846, "genuineness": 3.747946262359619, "identity": 0.18120869994163513, "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_NEWS_reg/6d6a6f12d496999666999a1315b60b3089f02292e8f04cf7a64ab700181da8bb.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": "6e7ef8ae9a05240b7d2da0a1488d5d2beac0f87e8b2bf36be66f292d862cf645", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4742fe925a8477ef75f7a8dd7b45df5e06911f753d9478450da72b74d8fe8d17", "generation_seconds_batch": 14.009174822014757, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6411268711090088, "target_z": 2.0937499290105164, "ce": 6.1020402908325195, "genuineness": 4.15588903427124, "identity": 0.6422158479690552, "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_NEWS_reg/6e7ef8ae9a05240b7d2da0a1488d5d2beac0f87e8b2bf36be66f292d862cf645.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "6ec3d1f3d8ca3c1e0aaee70dfc2f123f94d36bf439532db5e08503c10ef306aa", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8e5c7e5061a72e0223ae3030f571952fce03e7b2c17bc27c5e7c248a4f8a8e22", "generation_seconds_batch": 20.05922737705987, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8649725914001465, "target_z": 2.499999945912774, "ce": 6.165581703186035, "genuineness": 4.074300289154053, "identity": 0.17755377292633057, "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_NEWS_reg/6ec3d1f3d8ca3c1e0aaee70dfc2f123f94d36bf439532db5e08503c10ef306aa.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": "6ef3edc00aa23b278f08da4fa5d42e82c48b2234fde5ec1c6a57d2081c9cd2e4", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3f31ca533ab1434b2c964c7cab877d2f41d18523e2fcbbb4599cfb7cc56702f8", "generation_seconds_batch": 15.226651718025096, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4159774780273438, "target_z": 3.500000054087226, "ce": 6.205295085906982, "genuineness": 4.105680465698242, "identity": 0.6159214377403259, "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_NEWS_reg/6ef3edc00aa23b278f08da4fa5d42e82c48b2234fde5ec1c6a57d2081c9cd2e4.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": "6f3161885834bd522bb338431c7121b18147e4dcf93b56a5ec02ebcb4e4cd1e8", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Your message arrived after I had left.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7accb0811ce8bf42a6c43dce6be61073a803d5f8fbf719bf8e74ad4e4a626951", "generation_seconds_batch": 14.622531900997274, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1288645267486572, "target_z": 1.164062348724791, "ce": 6.324435234069824, "genuineness": 3.8420867919921875, "identity": 0.3521226644515991, "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_NEWS_reg/6f3161885834bd522bb338431c7121b18147e4dcf93b56a5ec02ebcb4e4cd1e8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "6f884efe6e36337e1a7ceb5cc191014a49a32723e03bb94c070c2075e9ff8748", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e11f119f4ee7c55ecf80ba185b685bcd8f41d45f586a7d89591824c01e6628d8", "generation_seconds_batch": 20.091951805981807, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.55503249168396, "target_z": 1.9375001555007736, "ce": 6.09409761428833, "genuineness": 4.319066047668457, "identity": 0.5912749767303467, "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_NEWS_reg/6f884efe6e36337e1a7ceb5cc191014a49a32723e03bb94c070c2075e9ff8748.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "6fa08ee02f5ed6aa04b954883503b20d3c157b4ab9f3c068eb895763c7fb2624", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "082a43b0dbae31ad504c8ad57571553a9b84749df4b6d310879b15c836fc1f1a", "generation_seconds_batch": 20.0458551060874, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.109982967376709, "genuineness": 3.9425034523010254, "identity": 0.5644947290420532, "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_NEWS_reg/6fa08ee02f5ed6aa04b954883503b20d3c157b4ab9f3c068eb895763c7fb2624.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "6fe83be7af2d547081f96746c68616102e9f6528e7682d5aa275daed2867b0ce", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3dddaf1b9ae121d0d1a69b3094f3a53d5f2b91a8e506dd2477219ca6ec666e1c", "generation_seconds_batch": 15.514612646074966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 6.181467056274414, "genuineness": 3.9676074981689453, "identity": 0.5821816921234131, "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_NEWS_reg/6fe83be7af2d547081f96746c68616102e9f6528e7682d5aa275daed2867b0ce.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "701afac096b51dbddc973f32c4db857d415d3426a29ab4a03ab5e75455ab2a5a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9ff784203c83de32986e8d25262ff170d165ebe1f36fe866fdffcd2e72f63422", "generation_seconds_batch": 14.93867017608136, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7016206383705139, "target_z": 0.3886718849300766, "ce": 6.2767791748046875, "genuineness": 3.9801597595214844, "identity": 0.4334936738014221, "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_NEWS_reg/701afac096b51dbddc973f32c4db857d415d3426a29ab4a03ab5e75455ab2a5a.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": "7055264cd92b8b5698dd924baa0d7ed14b69908b9078e6e44a1cfab6092c0b24", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "f867cb393d4d666f8d31149949691eccf8164f891b83e0b8b1aac56a91cd3943", "generation_seconds_batch": 16.028420988004655, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3570151329040527, "target_z": 1.5781251571909993, "ce": 6.284721851348877, "genuineness": 4.2751336097717285, "identity": 0.7221571207046509, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/7055264cd92b8b5698dd924baa0d7ed14b69908b9078e6e44a1cfab6092c0b24.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "707ec7b34e9527fbad115e8f5fbbce03aca4409da2ef0c014b12f2bccd45092f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b46a59adc07bab8625325b61ca8c513ef3c3eb51d574c2a975c0145dde0a4c7c", "generation_seconds_batch": 14.389329454046674, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6325175762176514, "target_z": 2.078125211278225, "ce": 5.839932441711426, "genuineness": 3.6789097785949707, "identity": 0.3664265275001526, "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_NEWS_reg/707ec7b34e9527fbad115e8f5fbbce03aca4409da2ef0c014b12f2bccd45092f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "709214c19f97d9f864fba7a1ee61ad8ea53c36935830f9313bdfb164433c1ce9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9b99ec234a504f91b497853981d3430e85d135cf62e7c4b6bc634e650250d461", "generation_seconds_batch": 14.884281608043239, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7186119556427002, "target_z": 2.234374984787968, "ce": 6.125868320465088, "genuineness": 4.168440818786621, "identity": 0.4701782763004303, "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_NEWS_reg/709214c19f97d9f864fba7a1ee61ad8ea53c36935830f9313bdfb164433c1ce9.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": "7152def3eb49b8b3c4ea3c6acac7333c62898a890b07fec7b0c30c459f5b6742", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "7ddd8ee5a01376bd81179f2a560c7a6d2f0781dd798fb8b3184532cd156a879e", "generation_seconds_batch": 19.323444967041723, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3957575559616089, "target_z": 1.6484374687308228, "ce": 6.205295085906982, "genuineness": 4.268857479095459, "identity": 0.5713752508163452, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/7152def3eb49b8b3c4ea3c6acac7333c62898a890b07fec7b0c30c459f5b6742.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "717f2b9c86ce8fcbc138270f98064a5eb7d07bc20f23af3c822a76e5b12440b5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a7925051522c75907332f2550a8147404326dedfb3e740c32548544120e678e2", "generation_seconds_batch": 21.301742951967753, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8821914196014404, "target_z": 2.531249814075162, "ce": 6.062326908111572, "genuineness": 4.105680465698242, "identity": 0.6587579250335693, "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_NEWS_reg/717f2b9c86ce8fcbc138270f98064a5eb7d07bc20f23af3c822a76e5b12440b5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "71f3b4c0c55a2a241c381e09c20bb85ecd940850573f55f4ca850f74d92652c2", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "644ec3d293f61301c11c9176a9f0c07365fb1959b8c9f33dc74e5db5386733da", "generation_seconds_batch": 14.757018235046417, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3570151329040527, "target_z": 1.5781251571909993, "ce": 6.4912309646606445, "genuineness": 4.287685871124268, "identity": 0.530957043170929, "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_NEWS_reg/71f3b4c0c55a2a241c381e09c20bb85ecd940850573f55f4ca850f74d92652c2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "72421fd20a00238c59c66a6103e7ec91c89e82175d94914d820ad8b4c5a2b3e3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "26770e3d241e79704861715956eeb0f3ac3e62f024c473a32c8905554d7fd971", "generation_seconds_batch": 19.50213654397521, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9510672092437744, "target_z": 2.656250152120322, "ce": 6.403861999511719, "genuineness": 4.124508857727051, "identity": 0.5000272393226624, "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_NEWS_reg/72421fd20a00238c59c66a6103e7ec91c89e82175d94914d820ad8b4c5a2b3e3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "72b2f351443b3ff1d5a69e728e47f86ddf9896238f8b8da35f6714e706688a45", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "0d46c9d2d703b5d5b5870035ed921ca3f939031c60e4b8398c21595e0dc54eb7", "generation_seconds_batch": 20.057255571940914, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5851655006408691, "target_z": 1.9921875329594032, "ce": 6.197352409362793, "genuineness": 4.143336772918701, "identity": 0.5713417530059814, "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_NEWS_reg/72b2f351443b3ff1d5a69e728e47f86ddf9896238f8b8da35f6714e706688a45.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "72c8f169b075950ac7d4b8916f5495b7bca567582888d77295e8c078cbcf19c1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cfb041c4ecc3659ec3cf551d5e13768d1d9a8bbb051d6927cd8f8a8ea779acd8", "generation_seconds_batch": 19.32386030408088, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2149591445922852, "target_z": 1.3203125549323385, "ce": 6.308549880981445, "genuineness": 4.306513786315918, "identity": 0.386576771736145, "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_NEWS_reg/72c8f169b075950ac7d4b8916f5495b7bca567582888d77295e8c078cbcf19c1.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": "72f36c6dbcf30c2a5de7c72bf6807822bfdba6dd664f0f539a4a501952addfbc", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6cd8b7acf2ab55140e9334822be9482871bbad794108487ebdd006650f4b4d03", "generation_seconds_batch": 14.243553403066471, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.0305562019348145, "genuineness": 4.030367851257324, "identity": 0.7651835680007935, "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_NEWS_reg/72f36c6dbcf30c2a5de7c72bf6807822bfdba6dd664f0f539a4a501952addfbc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "734fa83039618c848f90037ca93dba61c80089fc0b7658c2ec3f03e2e76f853e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f61f24a2ad81bbee3b99318540813180ec0cfe1d6ffdf5320c0c1d9aad602488", "generation_seconds_batch": 19.98474968795199, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.266615867614746, "target_z": 1.414062592117306, "ce": 6.173524379730225, "genuineness": 4.381826400756836, "identity": 0.5687382221221924, "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_NEWS_reg/734fa83039618c848f90037ca93dba61c80089fc0b7658c2ec3f03e2e76f853e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "7380276b72eb406e1a158b11baddc1d817b08a9b019cee00aacdbee0a29e3011", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4e840ea376861c3dfa6c7a89ad22a9438ec81a89376b66e93b7fdf2f56230b7a", "generation_seconds_batch": 20.150555234984495, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2795300483703613, "target_z": 1.437500101413548, "ce": 6.229123115539551, "genuineness": 4.325342178344727, "identity": 0.51959627866745, "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_NEWS_reg/7380276b72eb406e1a158b11baddc1d817b08a9b019cee00aacdbee0a29e3011.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "739c4a60a2cac2bc14407b5563cdd5722775ee2da923c579bee9d1a771c3c146", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d38f4ceb5d02578337ba841a7a8754a73e05589a811c5038f6c78986c5a4d630", "generation_seconds_batch": 21.66812719090376, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.873582124710083, "target_z": 2.5156250963428706, "ce": 6.38003396987915, "genuineness": 4.080576419830322, "identity": 0.5567678809165955, "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_NEWS_reg/739c4a60a2cac2bc14407b5563cdd5722775ee2da923c579bee9d1a771c3c146.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "744d7bba504b1f557d73f36d22648bd1e53f02a31c5b41e38484245abfde5524", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1456230c2ead117fb04d778ed7933bd5cc3536f5305d6cc2abb228884658a721", "generation_seconds_batch": 19.681391799938865, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1417787075042725, "target_z": 1.1874998580210328, "ce": 6.403861999511719, "genuineness": 4.519899368286133, "identity": 0.6388995051383972, "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_NEWS_reg/744d7bba504b1f557d73f36d22648bd1e53f02a31c5b41e38484245abfde5524.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "74c2e2d3ec0d3c1fc384b01f349fdf59b6505f5a41be1ba672e958bfc8ce75c6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "05939e92fb897488b7916d51a732e91fdb5b61efe3cf74c42c60c14945d49f78", "generation_seconds_batch": 15.164196260971949, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152985095977783, "target_z": 2.0468749104180324, "ce": 6.09409761428833, "genuineness": 4.068024158477783, "identity": 0.49888014793395996, "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_NEWS_reg/74c2e2d3ec0d3c1fc384b01f349fdf59b6505f5a41be1ba672e958bfc8ce75c6.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": "74d93829ef7967690f44671155ca378adf363eb12622c40dc0755be7bae94043", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1b8425bfc26257a06d455b65fed7ce21743fcb4842e0bf5f299ccd3346c4ab77", "generation_seconds_batch": 14.704422302078456, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0629899501800537, "target_z": 2.8593749442225485, "ce": 6.09409761428833, "genuineness": 4.017816066741943, "identity": 0.6127498149871826, "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_NEWS_reg/74d93829ef7967690f44671155ca378adf363eb12622c40dc0755be7bae94043.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "756c757c5afef09c2c3448d10e45ab33851c7db70c0241e1d30a0bb27ef39f31", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 0.5, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b22d8823692cd7d27578586f7ef44c747c0fedbedfba7fd528f3d3beb6d70e04", "generation_seconds_batch": 19.811923549044877, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3828433752059937, "target_z": 1.6249999594345808, "ce": 6.141753673553467, "genuineness": 4.406930446624756, "identity": 0.43584734201431274, "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_NEWS_reg/756c757c5afef09c2c3448d10e45ab33851c7db70c0241e1d30a0bb27ef39f31.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "75cb17470ecb96bf9ff14dc5da99a9ed17283f62718efcefe8751ee425478a25", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8f1c23be925aa30427d6c28bab8cb7da913ac078c6973165890b8268aeafa6b5", "generation_seconds_batch": 22.04508752701804, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1891307830810547, "target_z": 1.2734375363398547, "ce": 6.205295085906982, "genuineness": 4.632867813110352, "identity": 0.6898954510688782, "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_NEWS_reg/75cb17470ecb96bf9ff14dc5da99a9ed17283f62718efcefe8751ee425478a25.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "774256875d60b114d5f2a32c6cdf42bdd37deef409dfe4fcc9ab4a7480fa11e2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "bab3949202f512bc712a3825f46b8593783182096dfeb8e88a05d9b67390ac61", "generation_seconds_batch": 20.662725784932263, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7651153206825256, "target_z": 0.5039062335202984, "ce": 6.165581703186035, "genuineness": 4.532451152801514, "identity": 0.5789268016815186, "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_NEWS_reg/774256875d60b114d5f2a32c6cdf42bdd37deef409dfe4fcc9ab4a7480fa11e2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "775f06574a7644ecfc07ebb5fc01ae2364b31a2bcaf55f12e4ac9cd806f80f43", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2c539cb19fc8b39ace07b8b5553ee3894c813cda04a08fa2c55cc7feada7e5d2", "generation_seconds_batch": 18.767864911002107, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4401085674762726, "target_z": -0.0859375160571451, "ce": 6.340320587158203, "genuineness": 4.570107460021973, "identity": 0.5475398898124695, "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_NEWS_reg/775f06574a7644ecfc07ebb5fc01ae2364b31a2bcaf55f12e4ac9cd806f80f43.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "b951af673eacc817ff1540c9d26424fad0ae7dd8325cf6c335963da49abd544d", "generation_seconds_batch": 13.922070267028175, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1417787075042725, "target_z": 1.1874998580210328, "ce": 5.967014789581299, "genuineness": 4.381826400756836, "identity": 0.6121918559074402, "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_NEWS_reg/77962898fb95ddc0e92a153772191fe01482e40f98d2245158cb92652e8a93c1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "77bfa979dbe948210600a12bbd6e19d5a29494453f04d340ae858cb9789bed09", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ec7ade1e87fc542417c67813fd00e39fa45ccc12adfc0f5df5bd26ea51241144", "generation_seconds_batch": 22.236294489935972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4732426404953003, "target_z": 1.789062524508274, "ce": 6.038498878479004, "genuineness": 4.137060642242432, "identity": 0.5622310638427734, "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_NEWS_reg/77bfa979dbe948210600a12bbd6e19d5a29494453f04d340ae858cb9789bed09.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "77c88cd0f47aa1cd5b454bbb4373209775188ed7ab7fd1e10c1c61beb0ea5d94", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "79c07037e2b62947654be677e4301e155d039e3be8bc7f0e6eb12f415cdf9634", "generation_seconds_batch": 18.96763893100433, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9566755890846252, "target_z": 0.8515624771819517, "ce": 6.221180438995361, "genuineness": 4.469690799713135, "identity": 0.6983807682991028, "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_NEWS_reg/77c88cd0f47aa1cd5b454bbb4373209775188ed7ab7fd1e10c1c61beb0ea5d94.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "77e1892aaf8dc6f282b48d142a2062e4be2d28e3805aed04f2d82c32dad100fe", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "bd1a8720338a5afc7c550ddb88b73d8aac34dd593c51e0656455b6fb7fa098fe", "generation_seconds_batch": 14.884281608043239, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.140475034713745, "target_z": 3.0, "ce": 6.022613525390625, "genuineness": 4.036643981933594, "identity": 0.8154746294021606, "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_NEWS_reg/77e1892aaf8dc6f282b48d142a2062e4be2d28e3805aed04f2d82c32dad100fe.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "788f5ac207a10633cf4e457e53caca8e64d74e9117d84ed797058a9d3e063895", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "440a4d57a143da4fa6fe30fe9b3c6ccf5d8dea45f87eeda66507ed01c91a8bef", "generation_seconds_batch": 15.337763014016673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2537016868591309, "target_z": 1.3906250828210642, "ce": 6.23706579208374, "genuineness": 4.268857479095459, "identity": 0.7575287818908691, "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_NEWS_reg/788f5ac207a10633cf4e457e53caca8e64d74e9117d84ed797058a9d3e063895.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": "78eb43458b157ffc0970ba25f89c6c0ec02e9a3bf676012cbe7f854355b4bce0", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ce12588f86c45872c1e2f62855bb5cc88507b7618cd28f3695e9f061300fbe37", "generation_seconds_batch": 13.95950996899046, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 6.173524379730225, "genuineness": 4.118232727050781, "identity": 0.767625629901886, "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_NEWS_reg/78eb43458b157ffc0970ba25f89c6c0ec02e9a3bf676012cbe7f854355b4bce0.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": "79298fd103aba8f8dbb27c559d0351bed8e265ec267e9cd9d73ae77b5bb89787", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1a66fe8bf5079331eed8a90c32717b0030783feb572c985ba667a5561f11f045", "generation_seconds_batch": 14.463430935982615, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8477537631988525, "target_z": 2.468750077750387, "ce": 6.348263263702393, "genuineness": 4.143336772918701, "identity": 0.8495080471038818, "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_NEWS_reg/79298fd103aba8f8dbb27c559d0351bed8e265ec267e9cd9d73ae77b5bb89787.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "34b3b64ac32b34696d07658b17a34852d7d27c857c0435fe0552bfd16b7054d2", "generation_seconds_batch": 14.189571517985314, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3397960662841797, "target_z": 1.546874856330807, "ce": 6.133810997009277, "genuineness": 4.281409740447998, "identity": 0.3394971489906311, "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_NEWS_reg/794d3b9296e3c4be7c1e4182a2fa3cf095f7d593990763ce0f19c1f5e499a989.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "7ac8af40f188f8b81053c47a41934e48aa8f72f6ea3bf9339bf06fdc5337ab25", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "4042baa26e9e768fcba9e59094da8fb14bea6e5df27a3f89878df0adb945c0ce", "generation_seconds_batch": 14.26074289297685, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0729031562805176, "target_z": 1.0624999526736776, "ce": 6.324435234069824, "genuineness": 4.444586753845215, "identity": 0.46862077713012695, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/7ac8af40f188f8b81053c47a41934e48aa8f72f6ea3bf9339bf06fdc5337ab25.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "7b2b0a04bd537e1080284a06b4b8d1ecc43f9150d9f768b316ac84714ee833dc", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d7b9c72089668cacea7105a95c8e0e2ca7dbf0e72729ead772b04f932a1ee836", "generation_seconds_batch": 20.360407141968608, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1030362844467163, "target_z": 1.1171875464812095, "ce": 6.1894097328186035, "genuineness": 3.8169827461242676, "identity": 0.425057590007782, "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_NEWS_reg/7b2b0a04bd537e1080284a06b4b8d1ecc43f9150d9f768b316ac84714ee833dc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "7ba258171793115d328ed9c70306e5dbff3ace5630cc8ebbf69de13e5f230b9c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a58e69f31ebb71db3a580d1cd5dfdaccbf1fab499214ef0d350de0de57bc4607", "generation_seconds_batch": 15.222085061017424, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.167607069015503, "target_z": 1.2343748766135165, "ce": 6.4515180587768555, "genuineness": 4.21237325668335, "identity": 0.3498348593711853, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/7ba258171793115d328ed9c70306e5dbff3ace5630cc8ebbf69de13e5f230b9c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "7bc0f9e86bba298ba90957229181de5df85b58aed79c0e54ebabc464c8196842", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "980794924f859269890b211612b72573cf38dad22035ea4b8260b7219b27a793", "generation_seconds_batch": 19.098588880035095, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3699291944503784, "target_z": 1.601562450138339, "ce": 5.990842819213867, "genuineness": 4.068024158477783, "identity": 0.6586005687713623, "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_NEWS_reg/7bc0f9e86bba298ba90957229181de5df85b58aed79c0e54ebabc464c8196842.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "7bc2b52010b37d37ee6c27bdf44453844407c367999e524673da4be4d29d1c5a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f406881825acd935058fa47af4bec9f51ea10df8037fb98c049b65d1fec65f62", "generation_seconds_batch": 18.81211689300835, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6402782797813416, "target_z": 0.27734371577292755, "ce": 6.419747352600098, "genuineness": 4.532451152801514, "identity": 0.35870444774627686, "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_NEWS_reg/7bc2b52010b37d37ee6c27bdf44453844407c367999e524673da4be4d29d1c5a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "7c251c18a76f888ed90aaaf4d297c597b7b05bef6a83148e84c109712e6afa2e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e0275253afc3d680e76803c72f355c93320e0408dd537fbc9d5436e4afcaa792", "generation_seconds_batch": 18.736887493985705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5808608531951904, "target_z": 1.9843751740932574, "ce": 6.117925643920898, "genuineness": 4.149612903594971, "identity": 0.6288838386535645, "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_NEWS_reg/7c251c18a76f888ed90aaaf4d297c597b7b05bef6a83148e84c109712e6afa2e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "7ccbff69e60c87a1ad353ae64c4c04b34eaa6040f40032dcfbc90510e6d82036", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "00952617ff0d37fdaa0246a4b6c932477381c00b7bd5118d04a3becc7fc90b98", "generation_seconds_batch": 18.87042517401278, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7788782119750977, "target_z": 2.3437501724030314, "ce": 6.229123115539551, "genuineness": 4.18726921081543, "identity": 0.6781610250473022, "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_NEWS_reg/7ccbff69e60c87a1ad353ae64c4c04b34eaa6040f40032dcfbc90510e6d82036.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "7cdaa957f33f961422f14d30523a981834f10292050f12f97a623a098cc5ff04", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ff32077016182375a7f1fc4ed2cfa05a86ca74cee0caaec88a669044126a4fdb", "generation_seconds_batch": 19.005076337023638, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.150388240814209, "target_z": 1.203125008451129, "ce": 5.911416053771973, "genuineness": 4.457139015197754, "identity": 0.6097899079322815, "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_NEWS_reg/7cdaa957f33f961422f14d30523a981834f10292050f12f97a623a098cc5ff04.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "7d4f9818c85d6f9f885907d9404cbba7f7208d6aafcae9d0b4b9bfe9f523a770", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "498ac7e3d07fa30f69dcbceaa56eeffa47ad7b8330012ca494f142562d795d63", "generation_seconds_batch": 19.14715797593817, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0083322525024414, "target_z": 0.9453124061924681, "ce": 6.419747352600098, "genuineness": 4.256305694580078, "identity": 0.4374854564666748, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/7d4f9818c85d6f9f885907d9404cbba7f7208d6aafcae9d0b4b9bfe9f523a770.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "7dfb5d057d295c3dcb448605b07f90ef0acffa12097181bc6817771b7b132e80", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "dbbc10e32744c2b7e3326dfe5bff466f066132a518393d15286ba2ab06a3389b", "generation_seconds_batch": 18.703829491045326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0685985088348389, "target_z": 1.054687593807532, "ce": 6.165581703186035, "genuineness": 4.3127899169921875, "identity": 0.725980818271637, "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_NEWS_reg/7dfb5d057d295c3dcb448605b07f90ef0acffa12097181bc6817771b7b132e80.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "7e53174c5f97f823ae3ea68a0272a732637a9db410634f26c86c1060e47a37b9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f285376e96c607baec8270c0f1bda33523b0fb67114131b06e68960b6bf4eb57", "generation_seconds_batch": 19.085687605896965, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.054380416870117, "target_z": 2.8437497937924525, "ce": 6.109982967376709, "genuineness": 4.042920112609863, "identity": 0.6356814503669739, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het oude kraam. 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_NEWS_reg/7e53174c5f97f823ae3ea68a0272a732637a9db410634f26c86c1060e47a37b9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "7ebe2f3c5ddd202ed9fa46595c5013ef12a5d09ae30df1d7eac19705712f2d54", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "19632b8a02b9defaf1b0e9329f9058362115291558c620aa284b6d87b4929a9e", "generation_seconds_batch": 19.057467386941426, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.109982967376709, "genuineness": 4.218649387359619, "identity": 0.49071332812309265, "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_NEWS_reg/7ebe2f3c5ddd202ed9fa46595c5013ef12a5d09ae30df1d7eac19705712f2d54.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "00e89ba60d24351460b2ec8d3aac44e2fc82d37dbf2fae2dc360647ffa38f007", "generation_seconds_batch": 13.949739624047652, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8877999782562256, "target_z": 0.7265624636601453, "ce": 6.324435234069824, "genuineness": 4.061748504638672, "identity": 0.4165570139884949, "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_NEWS_reg/7f6fce1a6f733719f45a363f28fee4e24b2f07a3396aa07b42392913d7c24239.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "7fb9633d01bd99b063280adee146a16b5332e59f7612e2b2fd235c64c6f73927", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ffa93541364ec0ac1c55b23510921752a54760d70ad751969858a5c954c54e04", "generation_seconds_batch": 20.139590786071494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0715994834899902, "target_z": 2.875000094652645, "ce": 6.062326908111572, "genuineness": 4.20609712600708, "identity": 0.7431122064590454, "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_NEWS_reg/7fb9633d01bd99b063280adee146a16b5332e59f7612e2b2fd235c64c6f73927.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "7fdb81f2eddaa9fb6904c2dae20ab2519cca3282ec52f3b287e924236a5ceafc", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3ca5e58c908622527da5c2bf49003d17fa0f0dd485444b669f9405357d168c38", "generation_seconds_batch": 18.734744345070794, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.133169412612915, "target_z": 1.1718751402887415, "ce": 6.205295085906982, "genuineness": 4.432034492492676, "identity": 0.7041680812835693, "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_NEWS_reg/7fdb81f2eddaa9fb6904c2dae20ab2519cca3282ec52f3b287e924236a5ceafc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "806a5045648994f1fc6f7d88583cd116f4768d77271d2724266235c5312cf59e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "990352ab0270d52c44f2acc3981192412f1cd780682adcb45d2d1b509440f104", "generation_seconds_batch": 20.0458551060874, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3397960662841797, "target_z": 1.546874856330807, "ce": 6.141753673553467, "genuineness": 4.1998209953308105, "identity": 0.5147309303283691, "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_NEWS_reg/806a5045648994f1fc6f7d88583cd116f4768d77271d2724266235c5312cf59e.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "d8934a7ab85739b36abfface48c022d59fac78e2c43ed1c26ba52c6d81421638", "generation_seconds_batch": 15.458260831888765, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8060102462768555, "target_z": 0.5781250490165482, "ce": 6.260893821716309, "genuineness": 4.607763767242432, "identity": 0.47615480422973633, "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_NEWS_reg/80c2597a6b8f43b5338a6b37d654868d679f46ded48514b42837ab18fba327ad.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "81a5b1e3aac2f1ea53dc151eeb3cd09b61eb5b4b3fb4f403a6c087cd80409c66", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9ea39ded54fe93a01ec911107f57699fb74f186be4e7dc6522af47aa150c344a", "generation_seconds_batch": 18.787740224041045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4560236930847168, "target_z": 1.757812439996984, "ce": 5.95112943649292, "genuineness": 4.011539936065674, "identity": 0.694948673248291, "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_NEWS_reg/81a5b1e3aac2f1ea53dc151eeb3cd09b61eb5b4b3fb4f403a6c087cd80409c66.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "828b904e7f4b5a8c88c8bb25989724412dd10ee849e079b8b060a8aa37dac0f8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "483df7cfa99fc0ea6ef723050afad7e048fc2fe13fd450c2851e4decc9e85919", "generation_seconds_batch": 21.487543350085616, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5507276058197021, "target_z": 1.929687363936823, "ce": 6.141753673553467, "genuineness": 4.005263805389404, "identity": 0.5479012131690979, "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_NEWS_reg/828b904e7f4b5a8c88c8bb25989724412dd10ee849e079b8b060a8aa37dac0f8.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": "82c507d87ec8cf37e9707b5965fef68187bc133f1681f2a4dd4d5717cdc7cf65", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "b3e3679c2a372d957c7c85a3778bd1b40c8f2b2eb7fb66e552081605263bf82a", "generation_seconds_batch": 13.89549042400904, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.157639026641846, "genuineness": 4.218649387359619, "identity": 0.6211028695106506, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/82c507d87ec8cf37e9707b5965fef68187bc133f1681f2a4dd4d5717cdc7cf65.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "82d2fd0380d77efbade9c39f175cdaa7ba86fdb00682b09de513d1d461ad495b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7f7326711444b591c45c021843cd4a8474a9dca27a8469e040597c99b9d48689", "generation_seconds_batch": 20.47050705202855, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8426003456115723, "target_z": 0.6445311811232987, "ce": 6.268836498260498, "genuineness": 3.9299514293670654, "identity": 0.2992398738861084, "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_NEWS_reg/82d2fd0380d77efbade9c39f175cdaa7ba86fdb00682b09de513d1d461ad495b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "82f02a926a45514742b82e276b05b4fafc8c5ef44efe4ce0cf57b95f53f73a1b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "05ed8d32e2919132cc81b9e21147db409057bf447066278fc9bf883a54b1e794", "generation_seconds_batch": 20.707304392009974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 6.562715530395508, "genuineness": 4.15588903427124, "identity": 0.5042081475257874, "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_NEWS_reg/82f02a926a45514742b82e276b05b4fafc8c5ef44efe4ce0cf57b95f53f73a1b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "8301c2bc0a5bab2d1c043b6e91c9c6edb9778134f0cb9c09ca97226330f3fc4e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "79a7218168498192a55408d0dde250af3ee3780218fd31137d79655865c3844f", "generation_seconds_batch": 21.55270903604105, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5800121426582336, "target_z": 0.16796874450676616, "ce": 6.419747352600098, "genuineness": 4.482243061065674, "identity": 0.5072745084762573, "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_NEWS_reg/8301c2bc0a5bab2d1c043b6e91c9c6edb9778134f0cb9c09ca97226330f3fc4e.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": "837b55f71c1b2ea35c0c293f7e260acafd47440134a1ca6981a54bded1166307", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "e9e8be5880ac6c49d6646cdb61def2c78ea4836943924c3a6684dd283919b6da", "generation_seconds_batch": 13.920932088978589, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0371615886688232, "target_z": 2.812499925630065, "ce": 6.0305562019348145, "genuineness": 4.030367851257324, "identity": 0.5773586630821228, "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_NEWS_reg/837b55f71c1b2ea35c0c293f7e260acafd47440134a1ca6981a54bded1166307.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "8383261bea4102ea687620eedb73f47dbf22bc7a64646e24d35bfe387a4f1f5a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d8eb25fdf4f3f7430cda4034ad191db631fd7757ec3a6ac0541c16336e40006a", "generation_seconds_batch": 15.514612646074966, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7788782119750977, "target_z": 2.3437501724030314, "ce": 6.205295085906982, "genuineness": 4.005263805389404, "identity": 0.5203042030334473, "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_NEWS_reg/8383261bea4102ea687620eedb73f47dbf22bc7a64646e24d35bfe387a4f1f5a.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": "83adf294138b8f87763efe6a1a078e418c1670c05583974a28515e90572b9979", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "79c7aefdf17762f0935f8cf2430dc8b5906377147e93e1a98c97b896c2cf969c", "generation_seconds_batch": 20.16779401805252, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2278733253479004, "target_z": 1.3437500642285805, "ce": 6.23706579208374, "genuineness": 4.118232727050781, "identity": 0.7115365266799927, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/83adf294138b8f87763efe6a1a078e418c1670c05583974a28515e90572b9979.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": "842f56c34565afed0508ca4e96b88dcfcff38496b99db3261be43a3823a7fda6", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "17c920f21559c2b11691b6bcda73c899d1f0b464c57362fff169c80c7cdc892e", "generation_seconds_batch": 19.951172863016836, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.174912929534912, "target_z": 3.06250016902258, "ce": 6.332377910614014, "genuineness": 3.848362922668457, "identity": 0.5988962650299072, "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_NEWS_reg/842f56c34565afed0508ca4e96b88dcfcff38496b99db3261be43a3823a7fda6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "8481b414880c568c83040fc58041d70306ace18b900e39ded4c86983227c79e4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7db31c4db071f37a0871f622eee596843f4788ff653666a364647ae612371d53", "generation_seconds_batch": 18.883292009006254, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4086717367172241, "target_z": 1.6718749780270645, "ce": 6.181467056274414, "genuineness": 4.532451152801514, "identity": 0.8348247408866882, "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_NEWS_reg/8481b414880c568c83040fc58041d70306ace18b900e39ded4c86983227c79e4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "84ec11dbab59789fefd67265f8ce19b060ce2f05f98a51c37e6be52b8c6733c4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fc9ba8b717d716b902ed2f9a4ea6060b557e05b5bc6a71c36c0594f360f40c8b", "generation_seconds_batch": 21.407615242060274, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4775474071502686, "target_z": 1.7968750997233223, "ce": 5.990842819213867, "genuineness": 4.268857479095459, "identity": 0.6914173364639282, "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_NEWS_reg/84ec11dbab59789fefd67265f8ce19b060ce2f05f98a51c37e6be52b8c6733c4.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": "8576f869ec753bab1bbfaaa6fb237021c1ceed60ad3cdf832a5cde42f6488d33", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c43d39b304e25e6ffb65728134b395b54ce53080db598b76cb4ba12ca3e56ca1", "generation_seconds_batch": 14.949785306002013, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.85636305809021, "target_z": 2.4843747954826783, "ce": 6.340320587158203, "genuineness": 3.973883628845215, "identity": 0.4681895077228546, "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_NEWS_reg/8576f869ec753bab1bbfaaa6fb237021c1ceed60ad3cdf832a5cde42f6488d33.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "85b0e8289c231b0e34ff777f462beed4407804c3bcc75408fecad94cb4c36da8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e961a09544245ceafcd4464e8c56d1ca0d5f3eada2c4001feb905dbb62e2300c", "generation_seconds_batch": 15.474985713954084, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8649725914001465, "target_z": 2.499999945912774, "ce": 6.149696350097656, "genuineness": 4.162164688110352, "identity": 0.5149984359741211, "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_NEWS_reg/85b0e8289c231b0e34ff777f462beed4407804c3bcc75408fecad94cb4c36da8.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": "85bef42d60556570d0fff0d9f03801f729a6450f4f62acb4dd6a4b5c39b1ee36", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "64bad65dcd9c6638cbac2059ff24c136dafef5c01bfd3ae747935ed8c6c1dfef", "generation_seconds_batch": 20.025875468039885, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.260893821716309, "genuineness": 4.068024158477783, "identity": 0.752739429473877, "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_NEWS_reg/85bef42d60556570d0fff0d9f03801f729a6450f4f62acb4dd6a4b5c39b1ee36.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "864b7292aad11e6984f7cbb14cd319136025362a9894a3f1756b9c958c76ba57", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e3393f57eacecf40ea4cd3049de71e9b0f26ae4768ba457aa4b9920700a7943a", "generation_seconds_batch": 18.672109331935644, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6669552326202393, "target_z": 2.1406249476030004, "ce": 6.1894097328186035, "genuineness": 4.093128204345703, "identity": 0.7031831741333008, "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_NEWS_reg/864b7292aad11e6984f7cbb14cd319136025362a9894a3f1756b9c958c76ba57.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "86fe206f6963daa61040b8ae247ab4837980859c03e5d1c57a2cebff7fb86607", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3e7e6662aa9cf6efa410606c734dd5da13f204ea0900051372dde4c9e0d6da42", "generation_seconds_batch": 18.741132788010873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9166293144226074, "target_z": 2.593749983097742, "ce": 6.427689552307129, "genuineness": 4.20609712600708, "identity": 0.5084831714630127, "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_NEWS_reg/86fe206f6963daa61040b8ae247ab4837980859c03e5d1c57a2cebff7fb86607.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "8846749483c80e93c0dc1f3626091e4f685347a9b2b45552b4af7d84d3e8d1f4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cd340545e8a9a9b8142f971311645f0289bb9ce83271e726ce94cdcea7edb560", "generation_seconds_batch": 19.014288213918917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6209070682525635, "target_z": 0.24218756000301592, "ce": 6.55477237701416, "genuineness": 4.469690799713135, "identity": 0.29309409856796265, "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_NEWS_reg/8846749483c80e93c0dc1f3626091e4f685347a9b2b45552b4af7d84d3e8d1f4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "88488eef772b64fef08974a8beae7406e6d8c49304969e8c000686facdf3cfe9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "722cc86f65b101b50d7295c3e8088b1053b9605cf11ce01dcbdab67614441e9c", "generation_seconds_batch": 15.750117280054837, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1245598793029785, "target_z": 1.1562499898586451, "ce": 6.173524379730225, "genuineness": 4.337893962860107, "identity": 0.6159408688545227, "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_NEWS_reg/88488eef772b64fef08974a8beae7406e6d8c49304969e8c000686facdf3cfe9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "88873838645b60c3526bc1eaa6dfc5320f9f9d1f725176defdf560381fdaa669", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b8e2b90e2fb6b9624d22e9a6fc83f8e84fd044af409013459b33b2c18d86d9ac", "generation_seconds_batch": 19.679421687033027, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152985095977783, "target_z": 2.0468749104180324, "ce": 6.157639026641846, "genuineness": 4.337893962860107, "identity": 0.7145977020263672, "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_NEWS_reg/88873838645b60c3526bc1eaa6dfc5320f9f9d1f725176defdf560381fdaa669.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "88a985134eb7ca5810acd654eecff41337e05107b1c2998cedcd6a78a0e1079a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3efec93f01d947e0284f5bf144776b9ca7619479d04e92e0a85de8074da8a007", "generation_seconds_batch": 20.163828142918646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7651153206825256, "target_z": 0.5039062335202984, "ce": 6.507116317749023, "genuineness": 4.494794845581055, "identity": 0.4128528833389282, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/88a985134eb7ca5810acd654eecff41337e05107b1c2998cedcd6a78a0e1079a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "88b12430779cb0d5e3b5aad8f5df297293341e3d61d314d88122d306a36bec34", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4d6502e5fdec461862187aa6d6c2f316878cde5cb10385d5f13d2a52eaaa8989", "generation_seconds_batch": 15.40027757990174, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2795300483703613, "target_z": 1.437500101413548, "ce": 6.173524379730225, "genuineness": 4.406930446624756, "identity": 0.548168420791626, "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_NEWS_reg/88b12430779cb0d5e3b5aad8f5df297293341e3d61d314d88122d306a36bec34.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": "89102475e0af9e60405a1be40eb7b503ddae4071578ab67c8cc3ed4864adfda3", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ef040dcc931509d4d224b776a63c7e743cd7eff335b9224688dbaf20ce1e424a", "generation_seconds_batch": 20.21829507104121, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8305349349975586, "target_z": 2.4375002095879994, "ce": 6.260893821716309, "genuineness": 3.9927117824554443, "identity": 0.44263502955436707, "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_NEWS_reg/89102475e0af9e60405a1be40eb7b503ddae4071578ab67c8cc3ed4864adfda3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "8a0980745e52c51c27cad011c0eeaae7924d7955a78a307ad03d8bd4d9eb3205", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "305338610b30936bb061f04d5cbcf30018485d9a859f044ace3267d9a5cf7ea7", "generation_seconds_batch": 20.670553746982478, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 6.062326908111572, "genuineness": 4.080576419830322, "identity": 0.685019850730896, "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_NEWS_reg/8a0980745e52c51c27cad011c0eeaae7924d7955a78a307ad03d8bd4d9eb3205.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "8a45656827fb2605bd525ba7cdff21b12062a70651f57209452078182c036eb2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d145167757e010f26350093dc47972328062b87a9554597633f249f7696854de", "generation_seconds_batch": 18.741132788010873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.787487506866455, "target_z": 2.359374890135323, "ce": 6.419747352600098, "genuineness": 4.030367851257324, "identity": 0.2957551181316376, "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_NEWS_reg/8a45656827fb2605bd525ba7cdff21b12062a70651f57209452078182c036eb2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "8a5487d03eb5bb271cb55344b8809bfbe7ed01ae02b4eac693b5d5c3bcc06f42", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a9c4e9d1f6d1524af3686efbf9c84bcdaf321b06a58b1b46c4e9d6f444d40442", "generation_seconds_batch": 15.899023808073252, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3828433752059937, "target_z": 1.6249999594345808, "ce": 6.0305562019348145, "genuineness": 4.18099308013916, "identity": 0.654263973236084, "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_NEWS_reg/8a5487d03eb5bb271cb55344b8809bfbe7ed01ae02b4eac693b5d5c3bcc06f42.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "8b68f495f5ad6cf15fd5dc15a38537c96a1d9aec957857e8c5d5207f0f975d80", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d71b076b7961ddf1e65c69809306d628ba7b230ee036bd49150be551ae038a92", "generation_seconds_batch": 15.069717325037345, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4215859174728394, "target_z": 1.6953124873233065, "ce": 6.173524379730225, "genuineness": 4.243753433227539, "identity": 0.5941838026046753, "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_NEWS_reg/8b68f495f5ad6cf15fd5dc15a38537c96a1d9aec957857e8c5d5207f0f975d80.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "8c208c12f0048cda8d0f51106d76f5c7e4ca08d7b5360c6dc5ed31ca70d866be", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e5a596ea0cedc56350d504cefc62dfea12e693414233a31cb2c48cf27a209a16", "generation_seconds_batch": 21.06803124700673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5464229583740234, "target_z": 1.9218750050706774, "ce": 6.117925643920898, "genuineness": 4.369274139404297, "identity": 0.6242769360542297, "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_NEWS_reg/8c208c12f0048cda8d0f51106d76f5c7e4ca08d7b5360c6dc5ed31ca70d866be.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "8c3cdd9c2460db253bf0ea7d35142e73a33f837d728088e682346d4632f3ec60", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "755451a9b491c3738d1302adb4b26e3ccf6feb7ae7447b35d3c626a85ebaad02", "generation_seconds_batch": 14.724603501032107, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7013931274414062, "target_z": 2.2031251166255803, "ce": 6.213237762451172, "genuineness": 4.319066047668457, "identity": 0.6415844559669495, "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_NEWS_reg/8c3cdd9c2460db253bf0ea7d35142e73a33f837d728088e682346d4632f3ec60.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "8cb9332b7218694d23bd203be8b5c80b2b58413f206be347aa8ec4afc7e7f9af", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b6bbc5c09698767b65e1696a35b9db02175fa3c1a2668aed0b0d3a8e35348173", "generation_seconds_batch": 18.787740224041045, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 5.871703147888184, "genuineness": 4.086852550506592, "identity": 0.5993971824645996, "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_NEWS_reg/8cb9332b7218694d23bd203be8b5c80b2b58413f206be347aa8ec4afc7e7f9af.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "8cbafd2454d773fdfcba14dcaee532b941a47b42131f2455fae8fd2a63ce192e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "526e729ea226b8ce3e25e4dcd4e188364bdb4245e08d1b0e482d122490a12701", "generation_seconds_batch": 20.194974991027266, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5464229583740234, "target_z": 1.9218750050706774, "ce": 6.252951145172119, "genuineness": 4.306513786315918, "identity": 0.7342268228530884, "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_NEWS_reg/8cbafd2454d773fdfcba14dcaee532b941a47b42131f2455fae8fd2a63ce192e.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": "8cea129db59b1d8fdcd2d7ecc14dce0e82eb68f3d5357737c0fc14c2bb1103d8", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f2b53f53c547aeaa8d28f0367f393452da5898c5332794850da51187062a6203", "generation_seconds_batch": 19.83442033000756, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.174912929534912, "target_z": 3.06250016902258, "ce": 6.23706579208374, "genuineness": 4.218649387359619, "identity": 0.5838861465454102, "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_NEWS_reg/8cea129db59b1d8fdcd2d7ecc14dce0e82eb68f3d5357737c0fc14c2bb1103d8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "8cfd0a63b152bde4b3145bf74262d016d3273b5bfa7c7560dd57155607620be2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f83416b5b72283ff1c86d38750530eb627ad60d7dc017cac4ab03222f5e97cca", "generation_seconds_batch": 15.474985713954084, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.335491418838501, "target_z": 1.5390624974646614, "ce": 6.23706579208374, "genuineness": 4.20609712600708, "identity": 0.5961086750030518, "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_NEWS_reg/8cfd0a63b152bde4b3145bf74262d016d3273b5bfa7c7560dd57155607620be2.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": "8d4419d9f78520b24190cbf9a6b7fbadba5fbfaabe7e44ef7e812a501e2d9ff1", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I wrote the number on the envelope.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "befee6ae9066b73003e72dbc315303a575dd168267e968d50f77d7e672703368", "generation_seconds_batch": 14.66733354202006, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.352710247039795, "target_z": 1.5703123656270488, "ce": 6.2767791748046875, "genuineness": 3.810706615447998, "identity": 0.40632927417755127, "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_NEWS_reg/8d4419d9f78520b24190cbf9a6b7fbadba5fbfaabe7e44ef7e812a501e2d9ff1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "8dc45bab833693e0e9bc3e6b8d5c96ec2ae211e9927947e3cb07def5cc8d9cf3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "43126e985c95b6c25e707f63a361cce574380361d5a2efa39141a133365916d3", "generation_seconds_batch": 21.14479663304519, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.468937873840332, "target_z": 1.7812499492932259, "ce": 6.24500846862793, "genuineness": 4.13078498840332, "identity": 0.3782412111759186, "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_NEWS_reg/8dc45bab833693e0e9bc3e6b8d5c96ec2ae211e9927947e3cb07def5cc8d9cf3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "8dd2c1ad140b45a29e9e59d395995c8169fd8b80d91d6f9edc600e9b35078d2e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "714a328bd465bf3d70519ce489bf86ccbbe03410f8df605ce5b49ea20a2929e9", "generation_seconds_batch": 21.77064372599125, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.451719045639038, "target_z": 1.7500000811308385, "ce": 6.125868320465088, "genuineness": 4.18099308013916, "identity": 0.754298210144043, "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_NEWS_reg/8dd2c1ad140b45a29e9e59d395995c8169fd8b80d91d6f9edc600e9b35078d2e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "8ddbbdffc966a36768c14fbc4a6303bd062f55f29611092771426208b0f92e99", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f35ff476ac2d7408543270f8307f67471d7d6e2d12e8762e38202007225c8180", "generation_seconds_batch": 14.724603501032107, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6497364044189453, "target_z": 2.1093750794406128, "ce": 6.078212261199951, "genuineness": 4.042920112609863, "identity": 0.6501756906509399, "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_NEWS_reg/8ddbbdffc966a36768c14fbc4a6303bd062f55f29611092771426208b0f92e99.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "8e12c4547c5add43749566f711c9812a5035b833541b63721e85c144701f6070", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "eb5a7c09d4a60c17104b518a38f2362ab781bd21b7c1687e57f5b8d1fc13abd6", "generation_seconds_batch": 18.736887493985705, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 6.149696350097656, "genuineness": 4.105680465698242, "identity": 0.48920369148254395, "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_NEWS_reg/8e12c4547c5add43749566f711c9812a5035b833541b63721e85c144701f6070.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": "8e8f2addcb6c2b44725d680f3ce2b96e593a75927377869b6f2be1958687f17b", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "34987775a2d3cb970aed19a2910fd0c187de37618d23cd8cbe701eb7cf39200a", "generation_seconds_batch": 14.911016997066326, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.417281150817871, "target_z": 1.6874999121082583, "ce": 6.213237762451172, "genuineness": 4.15588903427124, "identity": 0.5241582989692688, "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_NEWS_reg/8e8f2addcb6c2b44725d680f3ce2b96e593a75927377869b6f2be1958687f17b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "8ea714e9783475dfa40b47c247f504af3a3a3dca9a998f2c158b76ffba92e8c4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "bca1c26639f540bca9974039b7042d9375fc24654ecf8b7bcbf80c8cf57b8bbc", "generation_seconds_batch": 20.986776006990112, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.088818311691284, "target_z": 2.9062499628150325, "ce": 6.1020402908325195, "genuineness": 3.9425034523010254, "identity": 0.6088159084320068, "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_NEWS_reg/8ea714e9783475dfa40b47c247f504af3a3a3dca9a998f2c158b76ffba92e8c4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "8f27ef8d09f8300796e7b5d54af361f4aa29e4bddcd4cbae08a175013675397a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "845267500089cb47f23392e4a689a72f51028f640b5a21bb30b87453d2469f4a", "generation_seconds_batch": 19.58714419102762, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 5.824047088623047, "genuineness": 4.080576419830322, "identity": 0.2896299362182617, "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_NEWS_reg/8f27ef8d09f8300796e7b5d54af361f4aa29e4bddcd4cbae08a175013675397a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "8f42f25ad0a0bca32056331c3c20450a51b531fc4737206c670df57734ad74f0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "1853bd4b3812456f1ec4f4f1fae3453915a7fa1a8b923a7c918eb3959a467432", "generation_seconds_batch": 18.87896600307431, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8426003456115723, "target_z": 0.6445311811232987, "ce": 6.213237762451172, "genuineness": 4.570107460021973, "identity": 0.7372882962226868, "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_NEWS_reg/8f42f25ad0a0bca32056331c3c20450a51b531fc4737206c670df57734ad74f0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "8f4ef332034b2a133598dfd8a7bf49cf0cc05f221abec2e0aa454a453dc49506", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "032ae4facfddaf907aa72613ddf4d4d58c8e96a516913bb92026c7323c393d8c", "generation_seconds_batch": 20.3575299680233, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1073410511016846, "target_z": 1.1250001216962575, "ce": 6.213237762451172, "genuineness": 4.507347106933594, "identity": 0.7058628797531128, "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_NEWS_reg/8f4ef332034b2a133598dfd8a7bf49cf0cc05f221abec2e0aa454a453dc49506.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "8f92e5a25e35e2f5df1e35ee59ca9bf2b04e438d8b7113a01416741e7eb34b9d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "74c5e9526b7e11bcb2bf7079673cb4e5df9ae6db0689cf903e5fab581c643d85", "generation_seconds_batch": 14.37828837998677, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2709205150604248, "target_z": 1.4218749509834518, "ce": 6.125868320465088, "genuineness": 4.287685871124268, "identity": 0.7256392240524292, "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_NEWS_reg/8f92e5a25e35e2f5df1e35ee59ca9bf2b04e438d8b7113a01416741e7eb34b9d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "90a5e7c7dec5099b3f539e26cdc7d87a934156442235a7dce3da4168c29c6e85", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f22cda7deab8b17a1c7892ad1e5045d2acda738465ffc49c7b46d29c4d153127", "generation_seconds_batch": 15.473690992104821, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8339909315109253, "target_z": 0.6289062470421048, "ce": 6.372091293334961, "genuineness": 4.381826400756836, "identity": 0.569172739982605, "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_NEWS_reg/90a5e7c7dec5099b3f539e26cdc7d87a934156442235a7dce3da4168c29c6e85.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "90c8c2867f71ce4e92046b7571723dc1eb7b56d104f696a39ec3c8b4cc95c169", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d005e2e33ff8e5f8073ecf23caacd6cbee90b7dfa5870f76c69feef8dde7ea23", "generation_seconds_batch": 14.27198091999162, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.787487506866455, "target_z": 2.359374890135323, "ce": 6.229123115539551, "genuineness": 4.419482707977295, "identity": 0.5814652442932129, "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_NEWS_reg/90c8c2867f71ce4e92046b7571723dc1eb7b56d104f696a39ec3c8b4cc95c169.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "9116c6847a13f690085c0ad87b607f2bf647fab854132cde18b367a1fc6ad03b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4062868e01df13ba042ec23588ca04eebbf7fb811af226d68d5fea16cb2d18aa", "generation_seconds_batch": 14.661677801981568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.890800952911377, "target_z": 2.546874964505258, "ce": 6.229123115539551, "genuineness": 4.256305694580078, "identity": 0.6419960260391235, "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_NEWS_reg/9116c6847a13f690085c0ad87b607f2bf647fab854132cde18b367a1fc6ad03b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "918bf2d94c8934331888d776c25f573a64b28e3b9e59cc966df35a2a25d4179f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "959bb0ee4f85a2b3bfc23cd0935a40afd29f3d653c11eeadeaea9e2703308925", "generation_seconds_batch": 19.018228149972856, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2493969202041626, "target_z": 1.382812507606016, "ce": 6.284721851348877, "genuineness": 4.2249250411987305, "identity": 0.4294421672821045, "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_NEWS_reg/918bf2d94c8934331888d776c25f573a64b28e3b9e59cc966df35a2a25d4179f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "919811cbf9c3c006819aa8da4691e5c55210f3ff16d8be71627f01a4c5bca1f8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "91ee587b3d5aab7a2cc63043b0769af79d35af91fa9debb7d3ea775da8ec8977", "generation_seconds_batch": 20.296078067971393, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.081512689590454, "target_z": 1.0781251031037737, "ce": 6.109982967376709, "genuineness": 4.375550270080566, "identity": 0.5828701853752136, "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_NEWS_reg/919811cbf9c3c006819aa8da4691e5c55210f3ff16d8be71627f01a4c5bca1f8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "91db2c735052f2287da94580f77f136e86ad6f27afd7e5a36bd60c766eb9617c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3db3a61959fb25f9410df639cd66e2de676deb747ac8b6412123e43c12e9d387", "generation_seconds_batch": 19.21775805298239, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.062326908111572, "genuineness": 3.9927117824554443, "identity": 0.43167224526405334, "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_NEWS_reg/91db2c735052f2287da94580f77f136e86ad6f27afd7e5a36bd60c766eb9617c.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": "921b3c7691bd9424fce9487d231ce67ffe4fb238fd90460e4ddf354a9e041277", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "3fb71881d8c09da10b02e057a1c5c053bb8e39ccbc79a886cfa460ce0c61d6b4", "generation_seconds_batch": 20.14756277098786, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.324435234069824, "genuineness": 4.15588903427124, "identity": 0.6766067743301392, "hypothesis": "Tu mensaje llegó después de que me fuera. Lo leí dos veces y guardé el teléfono. Dime cuándo podemos hablar.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/921b3c7691bd9424fce9487d231ce67ffe4fb238fd90460e4ddf354a9e041277.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "923e46d53417780f522093ade7ebce02fcb25845db1a7ba9c8390ed762ad1bcf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1c419e973e78debf4859bd31d95e1076792082a8a19f9b117f35f7c0746fa757", "generation_seconds_batch": 20.34802017000038, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.24500846862793, "genuineness": 4.111956596374512, "identity": 0.7741991877555847, "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_NEWS_reg/923e46d53417780f522093ade7ebce02fcb25845db1a7ba9c8390ed762ad1bcf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "9242ce25958b64b3cb614bc1b7479d3401e49b3f4b55739a4860f153531c531b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7752caaaa5d8329e2d8830d0ea8b7a57edd512951e700b271790e56e17db5829", "generation_seconds_batch": 19.24375373800285, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9609802961349487, "target_z": 0.8593749442225486, "ce": 6.213237762451172, "genuineness": 4.632867813110352, "identity": 0.7192234396934509, "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_NEWS_reg/9242ce25958b64b3cb614bc1b7479d3401e49b3f4b55739a4860f153531c531b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "929b417d1972f046238709caacd2f355dc528b3d0227c90d0166aafc073299ff", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2aa61863917cc8b78c6e3f424832e1b6d8be39c76eabc9f2aa39be53c91d830b", "generation_seconds_batch": 19.80228428193368, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4086717367172241, "target_z": 1.6718749780270645, "ce": 6.086154937744141, "genuineness": 4.287685871124268, "identity": 0.5131072402000427, "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_NEWS_reg/929b417d1972f046238709caacd2f355dc528b3d0227c90d0166aafc073299ff.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "92dfc887d77ba7923a7062c57182e41d9bfb3d26d2dd812f633b1a868f988636", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5ac2c1522e88dd01d62cd9c252787e7d1813e06b5f2073a2eb62d1b0959e2ee1", "generation_seconds_batch": 19.650221711955965, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.460328459739685, "target_z": 1.7656250152120323, "ce": 6.109982967376709, "genuineness": 3.8671910762786865, "identity": 0.532640814781189, "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_NEWS_reg/92dfc887d77ba7923a7062c57182e41d9bfb3d26d2dd812f633b1a868f988636.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "92e857ba487329e7520b4abf5ac2c01ecc8d4e3aaa355f4ba9ddd051424a1a25", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7dfb3e56edb8119c00071140b6e4b09dbdf9191ba9a8646b8857a4ccea688f02", "generation_seconds_batch": 18.803820917964913, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8133158683776855, "target_z": 2.4062499087278066, "ce": 6.229123115539551, "genuineness": 4.030367851257324, "identity": 0.5093674659729004, "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_NEWS_reg/92e857ba487329e7520b4abf5ac2c01ecc8d4e3aaa355f4ba9ddd051424a1a25.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "935de770f67a74a861cd3d836169539d264b99186df7798d49ceb9c9341ffd34", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "35b762e99fabbb7e9517382fa3456288f53d71a148680de94e277580a4bb3325", "generation_seconds_batch": 14.79409179405775, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1374740600585938, "target_z": 1.1796874991548871, "ce": 6.149696350097656, "genuineness": 4.394378662109375, "identity": 0.5393703579902649, "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_NEWS_reg/935de770f67a74a861cd3d836169539d264b99186df7798d49ceb9c9341ffd34.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "93de0e6c8c727bedcf67f7ab18b880c163d61fb747afd61aa6a9cdb062901466", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ae3fc9271a0c4fd3c1a301340dfb7778948f42e38ffac9b6e1da4e3cf5ef2dc8", "generation_seconds_batch": 15.591454515000805, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8791905641555786, "target_z": 0.7109375295789515, "ce": 6.23706579208374, "genuineness": 4.080576419830322, "identity": 0.2150391787290573, "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_NEWS_reg/93de0e6c8c727bedcf67f7ab18b880c163d61fb747afd61aa6a9cdb062901466.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "93ec5cc8f267e3e48857455e0eaee5fbeee907575bdb9d80b5fd880b2d25a21d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "eb7f379b93be29308632f278177fc33c76d14693a0d24295a6129f59cfddfe2a", "generation_seconds_batch": 20.502766470075585, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3785386085510254, "target_z": 1.6171873842195328, "ce": 5.982900142669678, "genuineness": 4.168440818786621, "identity": 0.5682320594787598, "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_NEWS_reg/93ec5cc8f267e3e48857455e0eaee5fbeee907575bdb9d80b5fd880b2d25a21d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "93f1168704a60dae15d00135731df3ebef63f70519ad90c9895adeffc48b660e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "fa8dbe94374195949e95131ba1402ab9c1a60bcf9020803663d97ad4529f66fc", "generation_seconds_batch": 19.41760154196527, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4000623226165771, "target_z": 1.6562500439458707, "ce": 6.09409761428833, "genuineness": 4.118232727050781, "identity": 0.6644009351730347, "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_NEWS_reg/93f1168704a60dae15d00135731df3ebef63f70519ad90c9895adeffc48b660e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "946657949a6c9ab4596d7cc417ff716f385a6a3c43c45cd04ab9e77e4b03e928", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "19c83e051bb197db7386449e4100a3001d535c5e6abf49ca2cbfb31c1aea1d41", "generation_seconds_batch": 20.98670075798873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.468937873840332, "target_z": 1.7812499492932259, "ce": 6.213237762451172, "genuineness": 4.519899368286133, "identity": 0.5703280568122864, "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_NEWS_reg/946657949a6c9ab4596d7cc417ff716f385a6a3c43c45cd04ab9e77e4b03e928.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": "94b2ff420c532a698608d398172f9167bfa6e2a2df3b61931f7ce215dae1c151", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I wrote the number on the envelope.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "c331d859a22da895ba00c2f66f4a334aa8c875349debb5d80ed6314d5b6cc2c9", "generation_seconds_batch": 21.272278312011622, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4990708827972412, "target_z": 1.8359373267518555, "ce": 6.292664527893066, "genuineness": 3.8797430992126465, "identity": 0.30071908235549927, "hypothesis": "I wrote the number on the envelope. Read it carefully and check the last two digits. Call me when you have a moment.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/94b2ff420c532a698608d398172f9167bfa6e2a2df3b61931f7ce215dae1c151.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "94b8316bb8b4be7802545cf6f03319a9108d93947b1e6df5d6b9dad87f7cd0b8", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "71a4a8c92554cc3959b76105d1e4fa919a516aaee671a82cf64db51e292557b2", "generation_seconds_batch": 15.833559063030407, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4474142789840698, "target_z": 1.7421875059157903, "ce": 6.292664527893066, "genuineness": 4.306513786315918, "identity": 0.7571951746940613, "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_NEWS_reg/94b8316bb8b4be7802545cf6f03319a9108d93947b1e6df5d6b9dad87f7cd0b8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "950713c1dccd347f03d80cb27e151208ba95094cb6ae058e65e4c0d3bd9bc925", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7928c2448438e6dc215b6d2595458a36b7c7422f4af1c293391e0a0172f6a388", "generation_seconds_batch": 18.801546472939663, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6841740608215332, "target_z": 2.171874815765388, "ce": 6.0305562019348145, "genuineness": 3.7918784618377686, "identity": 0.44139909744262695, "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_NEWS_reg/950713c1dccd347f03d80cb27e151208ba95094cb6ae058e65e4c0d3bd9bc925.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "9522a456dde2884bc3b70f65eacacc3b85ae5ecc5a3da9dcb44d8d6c9ca086e8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "606a68a154df109a7f4531b7f2ab4e6d5cc236ab569d0837b4b10fd7a0163566", "generation_seconds_batch": 14.620471233967692, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0018751621246338, "target_z": 0.9335936515443471, "ce": 6.427689552307129, "genuineness": 4.306513786315918, "identity": 0.373942106962204, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/9522a456dde2884bc3b70f65eacacc3b85ae5ecc5a3da9dcb44d8d6c9ca086e8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "957290ce949c3e61d8d38f45820c1db06ae51136cd90884dd4677484d376f547", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d0ac9b6e7fdb619fc5b750d6364297a494c1b1bf1e05f9fc274b144fb46c81b0", "generation_seconds_batch": 19.50213654397521, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 6.268836498260498, "genuineness": 4.218649387359619, "identity": 0.7898252010345459, "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_NEWS_reg/957290ce949c3e61d8d38f45820c1db06ae51136cd90884dd4677484d376f547.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "95919349b655949610f382c268e08a7c49b9eb16f044669848ad9364ed855620", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e3c0e8ad4ffbe4c637a7d9a69d8ba015f7f9eb94c0eca3c2500ef9c25b3a7844", "generation_seconds_batch": 19.565134981065057, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9855048656463623, "target_z": 2.7187498884450974, "ce": 5.967014789581299, "genuineness": 3.9676074981689453, "identity": 0.7739195823669434, "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_NEWS_reg/95919349b655949610f382c268e08a7c49b9eb16f044669848ad9364ed855620.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "959dc217232a3ebcd569f5a1eb82246028c73cc879b7d5d976c9537a58df1dad", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "74cdc84dec3b03e1cbe2b2681e2b0dd8bcf17a23516c3b3d9d432886420af0f5", "generation_seconds_batch": 18.56931138294749, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.965285062789917, "target_z": 0.8671875194375968, "ce": 6.38003396987915, "genuineness": 4.262581825256348, "identity": 0.35919541120529175, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/959dc217232a3ebcd569f5a1eb82246028c73cc879b7d5d976c9537a58df1dad.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": "95a28e77a9367027d32e26c5ebe83aeb9785dd82cc756bed07515450914f1a40", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I wrote the number on the envelope.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3bf21f69ee16f4d055f08cb3efb98d70f2746fa8e1668fb3977571fc9fcef6f2", "generation_seconds_batch": 21.272278312011622, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.38797664642334, "genuineness": 3.7542223930358887, "identity": 0.38822928071022034, "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_NEWS_reg/95a28e77a9367027d32e26c5ebe83aeb9785dd82cc756bed07515450914f1a40.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": "95cc282fcb1a6c3f92b75259f5a35730b801939739e84b6b56b7b654643c955d", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "38afed2a0a1ab63ee39356d3b2a9c6ad20ab03e9b2d3fad1aa4177532951d886", "generation_seconds_batch": 19.660428868955933, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.2767791748046875, "genuineness": 4.281409740447998, "identity": 0.7117588520050049, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/95cc282fcb1a6c3f92b75259f5a35730b801939739e84b6b56b7b654643c955d.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": "95f21136b15d5cbde341846dff7794d1abc38cf5995856a58e661b192efa91d5", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "573eab3e277754b58e5f1711a5bf3483d3923e10b4a195d8a25509bc65af19b0", "generation_seconds_batch": 15.1028476980282, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5119850635528564, "target_z": 1.8593748360480973, "ce": 6.221180438995361, "genuineness": 3.9801597595214844, "identity": 0.7157387733459473, "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_NEWS_reg/95f21136b15d5cbde341846dff7794d1abc38cf5995856a58e661b192efa91d5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "96137f774d4733bbaeb11f0b8d8746086afc035455d7473718cae169ae674bd6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "417d4cd1f5dff87999aa1a048bcc1835246a8f92aa83fd73b3532afab34a9ad1", "generation_seconds_batch": 18.87938659498468, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2537016868591309, "target_z": 1.3906250828210642, "ce": 6.173524379730225, "genuineness": 4.18099308013916, "identity": 0.6039882898330688, "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_NEWS_reg/96137f774d4733bbaeb11f0b8d8746086afc035455d7473718cae169ae674bd6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "96244c08c2990dde46f68003b31b23b90ed1f1e41cd60208d32260d34a42e0ad", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d816f06c9346413663cc36d42cf3d9b438b5be96c71ca79dc6f1abfa7472e5a1", "generation_seconds_batch": 19.122632598038763, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8821914196014404, "target_z": 2.531249814075162, "ce": 5.780362129211426, "genuineness": 3.8295347690582275, "identity": 0.6901755332946777, "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_NEWS_reg/96244c08c2990dde46f68003b31b23b90ed1f1e41cd60208d32260d34a42e0ad.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "963bdc9f043909c38e93d23cd5d3a22d3e30e50b5ddb47e82254260f1795ea6e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7d67001e54e8a79d707f24e975eab6d5bbae124c71d114173d79bf907ec6ec0b", "generation_seconds_batch": 19.571819069911726, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2364827394485474, "target_z": 1.3593749983097743, "ce": 6.308549880981445, "genuineness": 4.306513786315918, "identity": 0.5843538045883179, "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_NEWS_reg/963bdc9f043909c38e93d23cd5d3a22d3e30e50b5ddb47e82254260f1795ea6e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "96c53fd746399fae36a22fc8abc94812e477ca5e491f6f5ccecd50741c7ce5f0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "df55e857dca0942ff4db90eb31897a425e54bf58576a35acca547db90b712f30", "generation_seconds_batch": 15.872097639017738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7100024223327637, "target_z": 2.2187498343578715, "ce": 6.475345611572266, "genuineness": 4.124508857727051, "identity": 0.610939085483551, "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_NEWS_reg/96c53fd746399fae36a22fc8abc94812e477ca5e491f6f5ccecd50741c7ce5f0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "971fe48c51e0b0b3d74d79c1068af2e699e6b315ab062d6187ed05a0b54c9820", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "80ff7927acfb699480c7085e652d3a0eae7e51b19976b50a212c974f2d0750ac", "generation_seconds_batch": 18.902449395973235, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1245598793029785, "target_z": 1.1562499898586451, "ce": 6.1020402908325195, "genuineness": 4.494794845581055, "identity": 0.6709332466125488, "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_NEWS_reg/971fe48c51e0b0b3d74d79c1068af2e699e6b315ab062d6187ed05a0b54c9820.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "9773c0e760477c71e3b65c6c60a6d34a5d88ec636163a10ebfc72a9f8f022bf6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5dae9f07935fb92cc0ef36a8d1836d03a351aca0bb3883c12e79a5355f9d0dd3", "generation_seconds_batch": 15.287152311997488, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5076804161071777, "target_z": 1.8515624771819517, "ce": 6.141753673553467, "genuineness": 4.381826400756836, "identity": 0.5749372243881226, "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_NEWS_reg/9773c0e760477c71e3b65c6c60a6d34a5d88ec636163a10ebfc72a9f8f022bf6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "978110cf530fcab2d7a895c7e2e0aa533faa4d5a88dd2b06028badf80c8f570f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ac185040060cb7c2a572112450fd2c5b2f98093a91c8a999c64b7775aab29a2d", "generation_seconds_batch": 19.59390394098591, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6755647659301758, "target_z": 2.1562500980330963, "ce": 6.435632705688477, "genuineness": 4.105680465698242, "identity": 0.4743141829967499, "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_NEWS_reg/978110cf530fcab2d7a895c7e2e0aa533faa4d5a88dd2b06028badf80c8f570f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "978c9e09d95a104919b8a2817cc762fc87669778ebe1ec3a7b97524695da2a35", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7885979d49e638edf0fdbbc0c5a4ffd2c9416761065c232fcd4321fb4b8b26c8", "generation_seconds_batch": 21.122083690017462, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7199157476425171, "target_z": 0.421875059157903, "ce": 6.300607204437256, "genuineness": 4.061748504638672, "identity": 0.24693551659584045, "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_NEWS_reg/978c9e09d95a104919b8a2817cc762fc87669778ebe1ec3a7b97524695da2a35.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "97fe2ce9e42566490cf90a5695f34089c3a9e540f9244719cc9081750ffed708", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9ef29cef589349b061b6186be9f3290a88cf8c47757bc1dcb7418593877722dd", "generation_seconds_batch": 14.399691935046576, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1073410511016846, "target_z": 1.1250001216962575, "ce": 6.165581703186035, "genuineness": 3.8420867919921875, "identity": 0.1793018877506256, "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_NEWS_reg/97fe2ce9e42566490cf90a5695f34089c3a9e540f9244719cc9081750ffed708.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "986b52d0a3345e83f6f637f42093119ebf8c7b8d96d696d204f85a3c5208786c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f50619610c8c6df891f48a414435b6425c2f4e92c123a3fe0a268bbafb59d411", "generation_seconds_batch": 16.69160560402088, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5335087776184082, "target_z": 1.8984374957744354, "ce": 6.022613525390625, "genuineness": 4.193545341491699, "identity": 0.7784087061882019, "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_NEWS_reg/986b52d0a3345e83f6f637f42093119ebf8c7b8d96d696d204f85a3c5208786c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "98e3f8ad3fff9121da0c692dcdb4cfd6ad9adef26ffa94e7518497283c7d6f06", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "74bc2ac0466ee6f641e6ecfb49f4f816175ef09b33a4e31815e914831c28c061", "generation_seconds_batch": 18.81211689300835, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6887064576148987, "target_z": 0.36523437563383465, "ce": 6.356205940246582, "genuineness": 4.507347106933594, "identity": 0.31454065442085266, "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_NEWS_reg/98e3f8ad3fff9121da0c692dcdb4cfd6ad9adef26ffa94e7518497283c7d6f06.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "98e7f0a7780ae7bed359d99e8450e4459e631bde5ffaa652b77fdc545fe0db0d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "788cfb220f3f56b4b9b2e653e56f27e561694d8a8b634b6a79bc71a3e8c2c817", "generation_seconds_batch": 15.45976946305018, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.460328459739685, "target_z": 1.7656250152120323, "ce": 6.117925643920898, "genuineness": 4.281409740447998, "identity": 0.7045261263847351, "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_NEWS_reg/98e7f0a7780ae7bed359d99e8450e4459e631bde5ffaa652b77fdc545fe0db0d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "9915a692e5df03b277635511d3f163e517337f8e0dbcec9641e66ea9e9a7a9a1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "684c6beeeaac3ad84b4bd0654081d8fee478df47ea6bb3e7bcaf90d5db68a3df", "generation_seconds_batch": 15.80942339997273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1719118356704712, "target_z": 1.2421874518285647, "ce": 6.197352409362793, "genuineness": 4.444586753845215, "identity": 0.5136409997940063, "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_NEWS_reg/9915a692e5df03b277635511d3f163e517337f8e0dbcec9641e66ea9e9a7a9a1.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": "991c938cb97dc4d7c0516565f687999264af73baf04d3beff3d6ba5836121dd8", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Someone moved the little blue chair.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "421037f36b965f37b21df1344ac132ca3362271c10bb24d242a9c28e655769e4", "generation_seconds_batch": 20.91047477896791, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7616591453552246, "target_z": 2.312499871542839, "ce": 6.372091293334961, "genuineness": 4.13078498840332, "identity": 0.26125121116638184, "hypothesis": "Someone moved the little blue chair. It used to stand here, beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/991c938cb97dc4d7c0516565f687999264af73baf04d3beff3d6ba5836121dd8.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": "99a244775955e6a123f13255c5acc717c904d81cfef1f924dd78616b8431954a", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c0dd2afba5193e43c5f884cc0b2ca85f7066bab9a7d01f970f4c8d6051debbca", "generation_seconds_batch": 16.25254231202416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2364827394485474, "target_z": 1.3593749983097743, "ce": 6.499174118041992, "genuineness": 4.20609712600708, "identity": 0.6073230504989624, "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_NEWS_reg/99a244775955e6a123f13255c5acc717c904d81cfef1f924dd78616b8431954a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "99bdb0e11e8f516c47d10d52c11df2a540af49e8075ae6f076464d8097286020", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "906c6c32d3b3b8a81ea4d706be0354411b9901bd69889be0351b04769bd0dc9a", "generation_seconds_batch": 14.588781449012458, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4258906841278076, "target_z": 1.7031250625383545, "ce": 6.205295085906982, "genuineness": 4.507347106933594, "identity": 0.713726818561554, "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_NEWS_reg/99bdb0e11e8f516c47d10d52c11df2a540af49e8075ae6f076464d8097286020.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "39dc93b103a8c6e1f018afe97965590fd68faf0a0cb6417dac594abdff1cf5c3", "generation_seconds_batch": 15.010534388013184, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6166023015975952, "target_z": 0.2343749847879678, "ce": 6.284721851348877, "genuineness": 4.469690799713135, "identity": 0.4138396382331848, "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_NEWS_reg/99bf1b7d3eed7b885fdae888ab00b4960791c3971d259b46f771306d2801e39b.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": "99e5b24b44e131321c0f24fea41c0cc1d7bf8bf656301754026e629efe5b6ea1", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "5bbee922533f861f9f4cfbf88b2310f44bb559828e31ea5fb51221af86f758ca", "generation_seconds_batch": 14.379613516968675, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7530498504638672, "target_z": 2.296875153810548, "ce": 6.086154937744141, "genuineness": 4.287685871124268, "identity": 0.6686621308326721, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/99e5b24b44e131321c0f24fea41c0cc1d7bf8bf656301754026e629efe5b6ea1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "9af91d12746f858940b0d9690cb4e8914861c2cf05d0895e9da92c1470259d03", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d69128f84bf35c4bfc53c842b7f3f4d8544742e365df1a3648c4e0a8c99dfdd0", "generation_seconds_batch": 19.50242937298026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3311867713928223, "target_z": 1.5312501385985156, "ce": 6.086154937744141, "genuineness": 4.362998008728027, "identity": 0.784591555595398, "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_NEWS_reg/9af91d12746f858940b0d9690cb4e8914861c2cf05d0895e9da92c1470259d03.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "9b141d7059ea493a883e24b9627abebf0b97a48aa71dc1d2dcd57a102ce901f1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e62f7e971330ad612f2faa250eb355255b15644e339f3235edf6d520f3d97c25", "generation_seconds_batch": 18.803161473944783, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9338481426239014, "target_z": 2.62499985126013, "ce": 6.09409761428833, "genuineness": 3.9173994064331055, "identity": 0.42084044218063354, "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_NEWS_reg/9b141d7059ea493a883e24b9627abebf0b97a48aa71dc1d2dcd57a102ce901f1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "9b2cc3f0ca5280ce2b56af1a202de8c5ad3a526bf9bbdb1f498dd8d20965ef34", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cfe1f6f06439e0780c95f4eeb400f6aeed83ac3de7e11e46cd651593fd271405", "generation_seconds_batch": 19.12058468698524, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.717763364315033, "target_z": 0.4179687715503789, "ce": 6.4912309646606445, "genuineness": 4.432034492492676, "identity": 0.42126354575157166, "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_NEWS_reg/9b2cc3f0ca5280ce2b56af1a202de8c5ad3a526bf9bbdb1f498dd8d20965ef34.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "58b467cbf8d5eadbfaa99c4bd25f7e25f9cd19da479f5f0a87c1ba3c8100315d", "generation_seconds_batch": 15.062218261999078, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0027239322662354, "target_z": 2.7500001893052897, "ce": 6.284721851348877, "genuineness": 4.243753433227539, "identity": 0.77666175365448, "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_NEWS_reg/9c09e940c5dab53287ffae2d1a8fa5296f41b2df71714656aaa2bb8830bf5162.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "b4137d523b6e30c202c8d146c63de6d59e1bc12ea933c37b199dbe00df9df0d1", "generation_seconds_batch": 15.472013409947976, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8361432552337646, "target_z": 0.6328124264751777, "ce": 6.213237762451172, "genuineness": 4.6203155517578125, "identity": 0.6866525411605835, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al al otro lado.", "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_NEWS_reg/9c969db20a6db380e08592f6aed9c9189b3db634d677ebe8f573f593ebdfd2d3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "9caba66839425e29171a636bbbb9726169d2bb55d0470a7c4034a97049a08167", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "71010422896bde140f5369fe3ba45c8063b00e22dddfdc088a5e82569f7807c3", "generation_seconds_batch": 21.359934424981475, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.572251319885254, "target_z": 1.9687500236631612, "ce": 5.982900142669678, "genuineness": 3.8797430992126465, "identity": 0.5664782524108887, "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_NEWS_reg/9caba66839425e29171a636bbbb9726169d2bb55d0470a7c4034a97049a08167.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "9cb0788c4b469d633c9add9d69ca8620414fad71d4728657dc65c022e502626d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d573044b357fec12277298670bb3ffda2efccb0907afe7b2932c6faec9edb9bd", "generation_seconds_batch": 19.894581312895752, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.451719045639038, "target_z": 1.7500000811308385, "ce": 6.3641486167907715, "genuineness": 4.149612903594971, "identity": 0.5996629595756531, "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_NEWS_reg/9cb0788c4b469d633c9add9d69ca8620414fad71d4728657dc65c022e502626d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "9d3d02ffb5f21e3663617a683e42d11cca94ef106d4b1e9998ad18066201a685", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "8a41dd8b3bbc6662fdd53349ce34a031ecbd6c1e142303a3b9071d75f08e9364", "generation_seconds_batch": 19.681391799938865, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2149591445922852, "target_z": 1.3203125549323385, "ce": 6.24500846862793, "genuineness": 4.381826400756836, "identity": 0.738425612449646, "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_NEWS_reg/9d3d02ffb5f21e3663617a683e42d11cca94ef106d4b1e9998ad18066201a685.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "9d9f1c1c786f08e00f3f0e1c91c2896ad0527c4a1b1d552b06b963c7e84e98ca", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "1f3c0510d2ccf829bd06735c258f799a28a2d31af7f3a3a65bcf6ce0f6c2a936", "generation_seconds_batch": 14.02374214399606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.4540989398956299, "target_z": -0.06054686295714117, "ce": 6.427689552307129, "genuineness": 4.444586753845215, "identity": 0.44451525807380676, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/9d9f1c1c786f08e00f3f0e1c91c2896ad0527c4a1b1d552b06b963c7e84e98ca.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": "9defad751e2a94e3382b02d28f36a95e54dff00320b30a0ce9561e3c4b79930b", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "de521f321ce4fbc22ff443b444b769c582675a07e9699369afa986871cd5cf90", "generation_seconds_batch": 15.074431804940104, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.821925401687622, "target_z": 2.421875059157903, "ce": 6.09409761428833, "genuineness": 4.036643981933594, "identity": 0.7583668231964111, "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_NEWS_reg/9defad751e2a94e3382b02d28f36a95e54dff00320b30a0ce9561e3c4b79930b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "9e26240dd9dd1d5f3f45707664a4d69b89ecd2811a3ec31af6b2a437799dfe2f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "53d84cf8c08f9b59e0b8933f61b9232ee862a0890b9d9797f718af0ffe530a9a", "generation_seconds_batch": 19.05833646596875, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0642937421798706, "target_z": 1.0468750185924838, "ce": 6.022613525390625, "genuineness": 4.369274139404297, "identity": 0.5974916815757751, "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_NEWS_reg/9e26240dd9dd1d5f3f45707664a4d69b89ecd2811a3ec31af6b2a437799dfe2f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "9f3280e31e3c0641cc01c70624b5a0a8f268c631f4fed303ad0abef1126026d4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f00913351960c34c6d7974bd206fe4f76929594c5d9a58d10fe40b81acc15476", "generation_seconds_batch": 19.601447627996095, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.133169412612915, "target_z": 1.1718751402887415, "ce": 6.173524379730225, "genuineness": 4.400654315948486, "identity": 0.5761326551437378, "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_NEWS_reg/9f3280e31e3c0641cc01c70624b5a0a8f268c631f4fed303ad0abef1126026d4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "9fc9ed9e8d16c61d3aa65d3244ae44a57b2e4872fe2e95022b4ef46bbad3020a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "83a07aebf27ddbfa11aec189e2bba857c3f69cb8c86d1d2d8ec7f3c6edd72cbf", "generation_seconds_batch": 18.752185591030866, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 5.998785495758057, "genuineness": 4.068024158477783, "identity": 0.6388424038887024, "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_NEWS_reg/9fc9ed9e8d16c61d3aa65d3244ae44a57b2e4872fe2e95022b4ef46bbad3020a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "a033d0119487f9d7531e8e0c6c583d94260050fb35e815d1d95c25e175be8a39", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0b8c10fa2bcd83e5b581c6839a73d8c8f9ad62d4d19e5050ca27f9bd5e2fe7e0", "generation_seconds_batch": 18.803820917964913, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8133158683776855, "target_z": 2.4062499087278066, "ce": 6.308549880981445, "genuineness": 4.111956596374512, "identity": 0.6167992949485779, "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_NEWS_reg/a033d0119487f9d7531e8e0c6c583d94260050fb35e815d1d95c25e175be8a39.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": "a04cc45dbb689f368af49bf9c9d116c9acec89dbc5dbc4bce6d01212d26b2609", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I wrote the number on the envelope.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "33b37977d115b4eef0857fa5647eb976d0e384df8a73f193dab3c970c887e8e9", "generation_seconds_batch": 14.856012354022823, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3053584098815918, "target_z": 1.4843751200060318, "ce": 6.3641486167907715, "genuineness": 3.823258876800537, "identity": 0.49743250012397766, "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_NEWS_reg/a04cc45dbb689f368af49bf9c9d116c9acec89dbc5dbc4bce6d01212d26b2609.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "a103a770bfd69c237cd438f3c8b07d3dc1a552591a613d4f6d42d6405c94f3e3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cc00a02db370b5eb5eb2e4f793268ab93f7fe306b757cc2d344a79ad289d80e9", "generation_seconds_batch": 15.155631059082225, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0277035236358643, "target_z": 0.9804686701368309, "ce": 6.125868320465088, "genuineness": 4.557555198669434, "identity": 0.47668761014938354, "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_NEWS_reg/a103a770bfd69c237cd438f3c8b07d3dc1a552591a613d4f6d42d6405c94f3e3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "a10980ec055e6ef3a61a8027e1a600a6f6f3c74d01da5eb84efb67de12ea50e9", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8b55258b1e8a14196d6c614656fa7540dd319329ca51c9f1dda2ae5d3019306e", "generation_seconds_batch": 15.295344428042881, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2623111009597778, "target_z": 1.406250016902258, "ce": 6.252951145172119, "genuineness": 4.250029563903809, "identity": 0.3702351152896881, "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_NEWS_reg/a10980ec055e6ef3a61a8027e1a600a6f6f3c74d01da5eb84efb67de12ea50e9.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": "a174313777e01dab74d416ad356ee258726436964e0ee5a4de12e6e9163d29d6", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The keys are on the kitchen table.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9f546be3216f73548c35bd0f6bf0e8b7cc9176148c52ca592c6cc7945bf15f03", "generation_seconds_batch": 13.932612777105533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2278733253479004, "target_z": 1.3437500642285805, "ce": 6.435632705688477, "genuineness": 4.218649387359619, "identity": 0.3700976073741913, "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_NEWS_reg/a174313777e01dab74d416ad356ee258726436964e0ee5a4de12e6e9163d29d6.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": "a17a962d6d6205d2a2c7b5e29d508ffc651f43c95932a0ff9349cde025fb0ba9", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Your message arrived after I had left.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "66e85f2ad70c1e2f547686ac002e2fe88eaaec80fd211893f6e15c9ead84dfd9", "generation_seconds_batch": 20.496060185017996, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.851209819316864, "target_z": 0.6601562233789436, "ce": 6.395919322967529, "genuineness": 3.9927117824554443, "identity": 0.4359167218208313, "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_NEWS_reg/a17a962d6d6205d2a2c7b5e29d508ffc651f43c95932a0ff9349cde025fb0ba9.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": "a17e413f72193acbb1315ca0eec27b714136758a8edfd0409e02d61071931843", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "0d5ad1e94bf68c99d5d32fc8e7bc5d90a09b13c87ba28a997d5967804ce23f99", "generation_seconds_batch": 19.107649150071666, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3268818855285645, "target_z": 1.523437347034565, "ce": 6.181467056274414, "genuineness": 4.168440818786621, "identity": 0.6053897142410278, "hypothesis": "La carretera gira a la izquierda tras el puente. Sigue hasta la pequeña puerta de madera. Te esperaré al otro lado.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/a17e413f72193acbb1315ca0eec27b714136758a8edfd0409e02d61071931843.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": "a34989535cd2b61b8c883b83d6ab671358bedbc358c49479fc8bfc91feb6038e", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich schrieb die Nummer auf den Umschlag.\n(Natural clear speech.) [3.8 seconds duration] Lies sie genau... und prüfe die letzten beiden Ziffern.\n(Natural clear speech.) [3.2 seconds duration] Ruf an, wenn du Zeit hast.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "ccb6878c8cf2ad587444859e1cc8aa448359ddd7883ee6668a09bd34d5188a2f", "generation_seconds_batch": 13.974739265977405, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2093505859375, "target_z": 3.124999905347355, "ce": 6.38003396987915, "genuineness": 3.6663577556610107, "identity": 0.3368813991546631, "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_NEWS_reg/a34989535cd2b61b8c883b83d6ab671358bedbc358c49479fc8bfc91feb6038e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "a3e9676be7ca334dcfb5fa948f283c0f00916bc2b0d5a9b3fd84e94ecdc6d325", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "39f0d9d58d70cc193c8ff2f62eba925af2b06868c2b9960be870277309045947", "generation_seconds_batch": 19.54425967205316, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.165581703186035, "genuineness": 4.231201171875, "identity": 0.7096593379974365, "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_NEWS_reg/a3e9676be7ca334dcfb5fa948f283c0f00916bc2b0d5a9b3fd84e94ecdc6d325.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "a4179a313e0e9ff69d6e18add8eded35e578ef4e98ea421f10e52552e256ac21", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e19439f5195ddd7b356464ed0f737fbb1ecac13cc1b43329f5ba70d353d24406", "generation_seconds_batch": 19.05833646596875, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2364827394485474, "target_z": 1.3593749983097743, "ce": 6.038498878479004, "genuineness": 4.425758361816406, "identity": 0.7892276048660278, "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_NEWS_reg/a4179a313e0e9ff69d6e18add8eded35e578ef4e98ea421f10e52552e256ac21.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": "a462a5429a3144ea121bcca3100e5e5100e78875a0590a93d427e4d4c9e98472", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f442d3e8267bac83e02d4d16bdc746efdec76d31fe53e95641f2b21b07294e80", "generation_seconds_batch": 14.014309831894934, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.890800952911377, "target_z": 2.546874964505258, "ce": 6.141753673553467, "genuineness": 3.9550557136535645, "identity": 0.6837952733039856, "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_NEWS_reg/a462a5429a3144ea121bcca3100e5e5100e78875a0590a93d427e4d4c9e98472.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "a4dc994d1c83e3268bbf014d55aaafc23e7aa54bb6d537d2a12702bea28e955c", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "89926e13b178f2a8a45c3e1fc867d97200a7cfe26b5f5cbf3a70cedf9cd37e23", "generation_seconds_batch": 15.922158724046312, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0556843280792236, "target_z": 1.03125008451129, "ce": 6.252951145172119, "genuineness": 4.425758361816406, "identity": 0.7498029470443726, "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_NEWS_reg/a4dc994d1c83e3268bbf014d55aaafc23e7aa54bb6d537d2a12702bea28e955c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "a4f4823516821bcff710fc9e582cb31610eeefd9bc684d975acebb29a2427a01", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "21b69bf144d4d115226418c5e4508aa5a93cd4961e82719565c1a868b23de88c", "generation_seconds_batch": 14.920606210944243, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8305349349975586, "target_z": 2.4375002095879994, "ce": 6.395919322967529, "genuineness": 4.049196243286133, "identity": 0.5084065198898315, "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_NEWS_reg/a4f4823516821bcff710fc9e582cb31610eeefd9bc684d975acebb29a2427a01.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "a52b7925660c1e134b004df21b19970b78aea4e82139f3c1da465c586febef52", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dffd302ea1af702a6235e7d52d3c9943eb89d0c5683140a557ddbfaa7898008f", "generation_seconds_batch": 18.88741986604873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0685985088348389, "target_z": 1.054687593807532, "ce": 6.229123115539551, "genuineness": 4.18099308013916, "identity": 0.5539950132369995, "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_NEWS_reg/a52b7925660c1e134b004df21b19970b78aea4e82139f3c1da465c586febef52.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "a58930a9407b0c2bad339f0abd6d551f2d96a67c0654d041d87751dca6015452", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3ecee532bc71fe059a417f221f359b8dc15367d6f80433342f015274fa288840", "generation_seconds_batch": 19.71350035793148, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3225772380828857, "target_z": 1.5156249881684194, "ce": 6.221180438995361, "genuineness": 4.381826400756836, "identity": 0.5375781059265137, "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_NEWS_reg/a58930a9407b0c2bad339f0abd6d551f2d96a67c0654d041d87751dca6015452.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": "a5b9ba155526d4f8f984aa52000c6111d656c2f7f2f20e9e8a783e7fa640edef", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "c24a1ec60e44b8860d77dd1c52a44e6c08d7cb262a8a89a7be2886b4e62f276e", "generation_seconds_batch": 13.988149769953452, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6411268711090088, "target_z": 2.0937499290105164, "ce": 6.23706579208374, "genuineness": 4.218649387359619, "identity": 0.7481681704521179, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/a5b9ba155526d4f8f984aa52000c6111d656c2f7f2f20e9e8a783e7fa640edef.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": "a624f2d2b69e787235f952ba0e05062c1d194d67101ae92a51d1a4ee4b837945", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The keys are on the kitchen table.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "46d1efd6465009eeb6847546e46354ad0a933b6713dba162909b58364057a110", "generation_seconds_batch": 13.932612777105533, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0729031562805176, "target_z": 1.0624999526736776, "ce": 6.308549880981445, "genuineness": 4.394378662109375, "identity": 0.5540809035301208, "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_NEWS_reg/a624f2d2b69e787235f952ba0e05062c1d194d67101ae92a51d1a4ee4b837945.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "a6558a5fa59cd15f091aa9a6d292729166d35897f1739fb26af75f4b05c7ab4a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "26daea3ea0b31bbdc962835c776ae61964b41c298952b350668353e0d91b9a50", "generation_seconds_batch": 20.466894799028523, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0715994834899902, "target_z": 2.875000094652645, "ce": 6.483288764953613, "genuineness": 4.042920112609863, "identity": 0.45742714405059814, "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_NEWS_reg/a6558a5fa59cd15f091aa9a6d292729166d35897f1739fb26af75f4b05c7ab4a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "a676c86bad02359ba12b4d7ade1ba1aa88b8c4afd5d03a6de187f6c637d4c861", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8752b4c26e6234714cfc1dce78fa277c805a7de240b66c5a35a5517caa58a134", "generation_seconds_batch": 19.01794794294983, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8133158683776855, "target_z": 2.4062499087278066, "ce": 6.157639026641846, "genuineness": 4.111956596374512, "identity": 0.8689990639686584, "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_NEWS_reg/a676c86bad02359ba12b4d7ade1ba1aa88b8c4afd5d03a6de187f6c637d4c861.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "a6dbe9f9f832787e75a2e5fa0350d4b5dce1291a6eca856eec4f930298dfb584", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "f84a359dfae221ce01c6aa3cccc5d1f19437d4ef5233e7de70e4a340397d6612", "generation_seconds_batch": 20.502766470075585, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2795300483703613, "target_z": 1.437500101413548, "ce": 5.935244083404541, "genuineness": 4.319066047668457, "identity": 0.6448788642883301, "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_NEWS_reg/a6dbe9f9f832787e75a2e5fa0350d4b5dce1291a6eca856eec4f930298dfb584.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "a6fbec3de1b9cb1d2fd2fc15dadb6dcbb1454002836cf7c5d09789a4786bdae4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4f41554dc32c23696c565b4093ff6cd6f1e38d60889de61d46b395d42151d055", "generation_seconds_batch": 22.04508752701804, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152985095977783, "target_z": 2.0468749104180324, "ce": 6.2767791748046875, "genuineness": 4.331617832183838, "identity": 0.6657624244689941, "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_NEWS_reg/a6fbec3de1b9cb1d2fd2fc15dadb6dcbb1454002836cf7c5d09789a4786bdae4.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": "a7a8fc3331e6706ccd5122977eb09eedc2f53d6f0267339962c836f93430ac87", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "76d95e0f47a4c098420dccd13e4067178e29ef7233148560e55f71a67c9392db", "generation_seconds_batch": 15.15650177502539, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.829231023788452, "target_z": 4.249999918869162, "ce": 6.3641486167907715, "genuineness": 3.71028995513916, "identity": 0.44399821758270264, "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_NEWS_reg/a7a8fc3331e6706ccd5122977eb09eedc2f53d6f0267339962c836f93430ac87.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "a84757daa0e899f69db75c18c603a5dc29d057d58c9611ef27ead053aabe6cad", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.0, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5caf309085b6e5e766ceee8115f09a8115068a9e1a4d52260822d0d19d149c16", "generation_seconds_batch": 15.756918297032826, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.0146708488464355, "genuineness": 4.2249250411987305, "identity": 0.6211022138595581, "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_NEWS_reg/a84757daa0e899f69db75c18c603a5dc29d057d58c9611ef27ead053aabe6cad.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "a859b5a5369d2383ac44fadd39abf50e4325e0da7be88212a7526864517ddb89", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "c93e7f6d364e805ac35d5314ffb5c3d389c27b492e1c0d4032d3fdaaaec74ad8", "generation_seconds_batch": 15.222085061017424, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2450921535491943, "target_z": 1.374999932390968, "ce": 6.467403411865234, "genuineness": 4.218649387359619, "identity": 0.4208606481552124, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/a859b5a5369d2383ac44fadd39abf50e4325e0da7be88212a7526864517ddb89.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "a887452aa68edb8506c5e73ed3e0ea51ea6a88fb816d928345e7b9c8065880cd", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d253a0575d907a56425f61a85327e0ebaa49ba9160f1e569fe6616ea5d39c868", "generation_seconds_batch": 21.970107111963443, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.133169412612915, "target_z": 1.1718751402887415, "ce": 6.0305562019348145, "genuineness": 4.281409740447998, "identity": 0.6606652736663818, "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_NEWS_reg/a887452aa68edb8506c5e73ed3e0ea51ea6a88fb816d928345e7b9c8065880cd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "a88e7a04febf8999ec9489653a65c215ff67c182e21b2d0ad700cbec5216c367", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8df3acf25010932681569ebba4909420c7a95b5c25513138ce5036c95c02845e", "generation_seconds_batch": 20.48795889201574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3742339611053467, "target_z": 1.609375025353387, "ce": 6.133810997009277, "genuineness": 4.293961524963379, "identity": 0.6557272672653198, "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_NEWS_reg/a88e7a04febf8999ec9489653a65c215ff67c182e21b2d0ad700cbec5216c367.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": "a89a1709478abe7322451a13a5c69906acce7d9974cc53044dc9ea17aeca69db", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2ba8f193e0ada84b0e4caaa6d7e061b4789522ba3a0054c1edd9bf7ed22ae45e", "generation_seconds_batch": 20.14756277098786, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 6.157639026641846, "genuineness": 4.174716949462891, "identity": 0.700372576713562, "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_NEWS_reg/a89a1709478abe7322451a13a5c69906acce7d9974cc53044dc9ea17aeca69db.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "2b8a8623a25ece8fb0025dbcf3fb15b71d264489d424718191a02857596e609d", "generation_seconds_batch": 14.746553199947812, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3742339611053467, "target_z": 1.609375025353387, "ce": 6.006728172302246, "genuineness": 4.306513786315918, "identity": 0.5582125186920166, "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_NEWS_reg/a8bd001b1fcafd2f49b65d73e6ed12567e4c9664909a4d08700c9febfd977540.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "a8d7d3ebed841afb5de0a01abe1e1432108ac5ff9a7638cef5ffd7a43d2c813e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "670b4d5d66b0fdb8ee0ce0b56959232e129128e6cd16390dd9db8474021c7c5e", "generation_seconds_batch": 20.138614128925838, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6755647659301758, "target_z": 2.1562500980330963, "ce": 6.0305562019348145, "genuineness": 4.055472373962402, "identity": 0.7492941617965698, "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_NEWS_reg/a8d7d3ebed841afb5de0a01abe1e1432108ac5ff9a7638cef5ffd7a43d2c813e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "a8f1d85427c7617d512d0ce8b31ca1f7dff4af3cc33759e818d142bccc0f8bc3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "7e662458ad66261f70468d8c6b597901cacd470749bc0a8d0428fd6b96ce851d", "generation_seconds_batch": 20.088513071998022, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0642937421798706, "target_z": 1.0468750185924838, "ce": 6.356205940246582, "genuineness": 3.9173994064331055, "identity": 0.3115817904472351, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/a8f1d85427c7617d512d0ce8b31ca1f7dff4af3cc33759e818d142bccc0f8bc3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "a90176b81d307fc7d4533536272fba8c6f666c010817cf59b0bafd94b9d53df0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ec74853cab772d546ffb883567e2eb9e46689236c4b28163044e7a9bb7a85f02", "generation_seconds_batch": 18.999166651046835, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.780181884765625, "target_z": 0.5312500304240644, "ce": 6.38003396987915, "genuineness": 4.419482707977295, "identity": 0.3297390639781952, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/a90176b81d307fc7d4533536272fba8c6f666c010817cf59b0bafd94b9d53df0.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "e7e353cda38763ece05dc4f208c60a214cfcb293b9ba3886e1d0024e6ed5246f", "generation_seconds_batch": 14.116779235075228, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.666106641292572, "target_z": 0.3242187343654114, "ce": 6.292664527893066, "genuineness": 4.632867813110352, "identity": 0.6306825876235962, "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_NEWS_reg/a94b7d5da78ff217f8e5a4f7ae4520900bef0ad764d0e4ee898e69de10c28aea.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "a94e065dbc43d7d590cf510ea56fa5304d3e7408061767e348892c11a9fbee88", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5ac6d706cace46546c2d59a51cabff9d168220e380d706869a89e96d194456ef", "generation_seconds_batch": 18.734575642971322, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8490574359893799, "target_z": 0.6562499357714195, "ce": 6.1894097328186035, "genuineness": 3.9927117824554443, "identity": 0.27460771799087524, "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_NEWS_reg/a94e065dbc43d7d590cf510ea56fa5304d3e7408061767e348892c11a9fbee88.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "a9b615b37772df8220d7d91b3a75e05ae78c00a720862dd152603ea3d1fac253", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e0f4491ffc18eb18a5b6b8bf372baeedb6191bd480d5d69b3e908c3f2abd3cea", "generation_seconds_batch": 20.32046227401588, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7530498504638672, "target_z": 2.296875153810548, "ce": 6.403861999511719, "genuineness": 4.250029563903809, "identity": 0.4972730875015259, "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_NEWS_reg/a9b615b37772df8220d7d91b3a75e05ae78c00a720862dd152603ea3d1fac253.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": "a9ea7d5635967978619926932c4b5d3f30724b07aa972651cc483dc1029554a9", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "73145241513ec0620aea9c4c2ca5cce82fa8e92df5f8aa2d37c03c5fd82d72f9", "generation_seconds_batch": 19.61420097493101, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6841740608215332, "target_z": 2.171874815765388, "ce": 6.070269584655762, "genuineness": 4.030367851257324, "identity": 0.43201807141304016, "hypothesis": "De weg gaat linksaf na de brug, loop verder, tot aan het kleine Houtenhek. Ik wacht aan de andere kant.", "wer": 0.09523809523809523, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/a9ea7d5635967978619926932c4b5d3f30724b07aa972651cc483dc1029554a9.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": "ab033ea6f6652fbdb1c59d9f21ebadffafd69e629fc1261114f9c96a8510e5e4", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0779d6d2299a6392a308125ff1b67c1dec1f3cc750849a41cee15613af674e48", "generation_seconds_batch": 14.233767705969512, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1073410511016846, "target_z": 1.1250001216962575, "ce": 6.3641486167907715, "genuineness": 4.231201171875, "identity": 0.4930979609489441, "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_NEWS_reg/ab033ea6f6652fbdb1c59d9f21ebadffafd69e629fc1261114f9c96a8510e5e4.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": "ab62f5d63a9b304b06a04c3326210319a74d0bde77b52caac587c58b2b7e273c", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Someone moved the little blue chair.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "37977663a4173a1577efe3bcdcdcc7c3b2a9862db6378b1f90d841fd4bf4ff5f", "generation_seconds_batch": 14.26039441791363, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6239080429077148, "target_z": 2.0625000608481288, "ce": 6.459460258483887, "genuineness": 4.193545341491699, "identity": 0.3614749610424042, "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_NEWS_reg/ab62f5d63a9b304b06a04c3326210319a74d0bde77b52caac587c58b2b7e273c.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 127, "seconds_generated": 10.16, "mp3_sha256": "5b2342857ff77fa4b3bc8570ee03f002e7a0b3f34296fdf66d2e9d7a0eb5a310", "generation_seconds_batch": 13.868217114009894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6219832301139832, "target_z": 0.24414064971955232, "ce": 6.475345611572266, "genuineness": 4.325342178344727, "identity": 0.3173861503601074, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/ab7a94299f968552a7b966fa3391724b652d8e20483888a7767196f29a7f4a69.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "ac8867ce490eac6709a908228a679ab1449bb212c263be77231b5fa0323b3a42", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "81ad961d1eaed7e6ecabbb655e8d76993316c2ec606c580726903696305c62ad", "generation_seconds_batch": 20.330752862035297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4904615879058838, "target_z": 1.820312609019564, "ce": 6.141753673553467, "genuineness": 4.055472373962402, "identity": 0.5768417716026306, "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_NEWS_reg/ac8867ce490eac6709a908228a679ab1449bb212c263be77231b5fa0323b3a42.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "aca0eafdd077066c8a594f46cb055735c14599aeda3e65d5967bd8871fd0795f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "962b5cd76d90c2a45b22ae81d1bec9c10cc9cc308660985ce97c8c834c123b1c", "generation_seconds_batch": 18.7434428720735, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9157807230949402, "target_z": 0.7773437698601532, "ce": 6.221180438995361, "genuineness": 4.042920112609863, "identity": 0.22353166341781616, "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_NEWS_reg/aca0eafdd077066c8a594f46cb055735c14599aeda3e65d5967bd8871fd0795f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "ad0304417f8f54d16b4c7b17b0402fa7ef936bb131ed4be0ea56b429ffec5a2e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2e62a53a6d7f815239637bdfd18df11e0cbbeebe2fb645d7cf62ee97fc84cbeb", "generation_seconds_batch": 14.307129143038765, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6800969839096069, "target_z": 0.34960933337818967, "ce": 6.109982967376709, "genuineness": 4.457139015197754, "identity": 0.45431843400001526, "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_NEWS_reg/ad0304417f8f54d16b4c7b17b0402fa7ef936bb131ed4be0ea56b429ffec5a2e.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "936b0905a57d24f2c044b89ebbdb191c2fe48ce57de789bab66369235ac0c8a8", "generation_seconds_batch": 13.968352608033456, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5708646178245544, "target_z": 0.15136721148007853, "ce": 6.038498878479004, "genuineness": 4.457139015197754, "identity": 0.5409024953842163, "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_NEWS_reg/adcc0121d6c11de6953f8243f84ad890832f75283c96342c1db246fddc443438.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "ae24aca5209a9298cb40f8075b1de7ec4ddc01edef54f4c3ea53eb8770912fb1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "047c00008c6b18d016681a7c85918d2504773a337bf8aabe5f4b467db036404a", "generation_seconds_batch": 15.392544595059007, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8985618352890015, "target_z": 0.7460937935233144, "ce": 6.221180438995361, "genuineness": 3.835810661315918, "identity": 0.3166775703430176, "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_NEWS_reg/ae24aca5209a9298cb40f8075b1de7ec4ddc01edef54f4c3ea53eb8770912fb1.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "2402fedb636509e6428aabfeb2101974560a08696687e41ca95d291d11ec03ef", "generation_seconds_batch": 13.991360471001826, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9695897698402405, "target_z": 0.8749999864781935, "ce": 6.435632705688477, "genuineness": 4.143336772918701, "identity": 0.28474313020706177, "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_NEWS_reg/ae3130a5db36c2790a31e23135fde257111c56fa9c5b8061c1f3e51a8cb61258.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": "aea1506fa0df17206baa94739d615458aefc15a4d58f4200263a13734925b70b", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wann wir reden können.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Natural clear speech.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wann wir reden können.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "89ffea732f568a47f7915fc63c518853a1b992deb0a74f8e319af751d775b9c1", "generation_seconds_batch": 16.25254231202416, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0858173370361328, "target_z": 1.0859374619699196, "ce": 6.340320587158203, "genuineness": 4.174716949462891, "identity": 0.3824186325073242, "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_NEWS_reg/aea1506fa0df17206baa94739d615458aefc15a4d58f4200263a13734925b70b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "aea239d3f2115727bef9aee4620a6ac08d85f0624e43e6e504d4abfda74ddb73", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3bbdfc2be929a8de7978da8473dc61783e2adcfcacf30821357b0ac9536ab9dc", "generation_seconds_batch": 18.827360568917356, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0298559665679932, "target_z": 0.9843750659188062, "ce": 6.157639026641846, "genuineness": 4.545003414154053, "identity": 0.6912013292312622, "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_NEWS_reg/aea239d3f2115727bef9aee4620a6ac08d85f0624e43e6e504d4abfda74ddb73.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "af136051944264d73f2bb2d64aee902e0e8f965294933897a8adea5f0ebbef92", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5507bf8d7bc7be0865638e2c138dacdba23db38dd5b7ca0e8048dbe451ace09b", "generation_seconds_batch": 15.80942339997273, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.109982967376709, "genuineness": 4.394378662109375, "identity": 0.7131708860397339, "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_NEWS_reg/af136051944264d73f2bb2d64aee902e0e8f965294933897a8adea5f0ebbef92.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "af2e79304ad54c986d05895f21db34183352a9e24c24e2bacd0d52dcc6eea3cc", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f39f921e8144127edd0a1cc44e60ed6810d444973b71909632702245041c39e6", "generation_seconds_batch": 21.06803124700673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5076804161071777, "target_z": 1.8515624771819517, "ce": 5.982900142669678, "genuineness": 4.042920112609863, "identity": 0.7474784255027771, "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_NEWS_reg/af2e79304ad54c986d05895f21db34183352a9e24c24e2bacd0d52dcc6eea3cc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "afd2a6ea86577f2e377c7188330a9369ed640c58e91939019a3d647aae3687cd", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e3dc7f4648f874378f4b0c9fa3b8c5009fb35cea17bb4302a9d0a2ea6ceced9d", "generation_seconds_batch": 19.24375373800285, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8253815174102783, "target_z": 0.613281312960911, "ce": 6.24500846862793, "genuineness": 4.545003414154053, "identity": 0.6456630229949951, "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_NEWS_reg/afd2a6ea86577f2e377c7188330a9369ed640c58e91939019a3d647aae3687cd.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "afdebecad073e172188212f85bb2ed00ddc7939b4dcf2eace7a3a58f0a38313f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f96f4592f5f4e6866b17f08d86832bcb693a5ef34de0ede5370d2d8f496099c9", "generation_seconds_batch": 21.514967605937272, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4000623226165771, "target_z": 1.6562500439458707, "ce": 6.229123115539551, "genuineness": 4.419482707977295, "identity": 0.6019822359085083, "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_NEWS_reg/afdebecad073e172188212f85bb2ed00ddc7939b4dcf2eace7a3a58f0a38313f.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": "b0208c616368b918f92e37707075fd4c0db5883db24edd57e6cef8a67030cc1e", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We zetten het na de lunch terug.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Iemand heeft het kleine blauwe stoeltje verplaatst.\n(Natural clear speech.) [3.8 seconds duration] Het stond hier... naast de hoge boekenkast.\n(Natural clear speech.) [3.2 seconds duration] We zetten het na de lunch terug.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6fd6a8e0810ac2f307398eae50ca422becd32b6c9bf4cd7a206a316819b4aa73", "generation_seconds_batch": 15.1028476980282, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8821914196014404, "target_z": 2.531249814075162, "ce": 6.284721851348877, "genuineness": 4.074300289154053, "identity": 0.4574328362941742, "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_NEWS_reg/b0208c616368b918f92e37707075fd4c0db5883db24edd57e6cef8a67030cc1e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "b03f12e75aa894f10aba67420f4e1711ed614a36a5f92010a34219a1b3fce51f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "10a77e4c5e14c066a731f9b5c4c8f3a8a74db846974a7b403abdc225040acab7", "generation_seconds_batch": 19.0654991309857, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5980796813964844, "target_z": 2.015625042255645, "ce": 6.125868320465088, "genuineness": 4.369274139404297, "identity": 0.6225448846817017, "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_NEWS_reg/b03f12e75aa894f10aba67420f4e1711ed614a36a5f92010a34219a1b3fce51f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "b059d5b664ba65262fa931ef4e80ce11d5ce9961d717cf0093b66c3b34983102", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e5227a22f05ed7d02ea2d07a670cc93368233a5111bbbbed6a97b39cb6b4ed11", "generation_seconds_batch": 15.839115922921337, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4560236930847168, "target_z": 1.757812439996984, "ce": 6.109982967376709, "genuineness": 4.168440818786621, "identity": 0.5999702215194702, "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_NEWS_reg/b059d5b664ba65262fa931ef4e80ce11d5ce9961d717cf0093b66c3b34983102.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "b0bdafd190bfb7732772d1c65f96ca62a74c606dde95dc13aa5bf1c93d8afcd7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5f0a8cf817131b5f0beda9fb4e1d18ba7dffbe8363d1143ddbc9fc411be376d1", "generation_seconds_batch": 19.41760154196527, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8133158683776855, "target_z": 2.4062499087278066, "ce": 6.23706579208374, "genuineness": 4.21237325668335, "identity": 0.6523423194885254, "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_NEWS_reg/b0bdafd190bfb7732772d1c65f96ca62a74c606dde95dc13aa5bf1c93d8afcd7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "b0e1d25a9526052f4703ae9caebfa7b190978a53c04bcb6a8041bbce833ce777", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1a65775bd3a25594e0ac6c87c4fc0941ed5d599a9827005417d5e2a5f7eef154", "generation_seconds_batch": 14.891127925016917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.468937873840332, "target_z": 1.7812499492932259, "ce": 6.046441555023193, "genuineness": 4.149612903594971, "identity": 0.48077210783958435, "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_NEWS_reg/b0e1d25a9526052f4703ae9caebfa7b190978a53c04bcb6a8041bbce833ce777.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "b12eadeb5022d821da76ebe53f157f2fe1f67f593196b7435a50d1e095c7bd09", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5390b6f6cfeb3a109575a858071389ac516e3b2b55a600ec8220807799c1ed82", "generation_seconds_batch": 16.310151770012453, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7100024223327637, "target_z": 2.2187498343578715, "ce": 6.173524379730225, "genuineness": 4.381826400756836, "identity": 0.6324124932289124, "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_NEWS_reg/b12eadeb5022d821da76ebe53f157f2fe1f67f593196b7435a50d1e095c7bd09.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "b17518759e7ba06eca72aace63d85df00156a75085c990eab244efcc7499e903", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3fbd31b822c7fa057b8025d61ea1643809793f3cf9bc4d89de1f8c446300d486", "generation_seconds_batch": 14.661677801981568, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.174912929534912, "target_z": 3.06250016902258, "ce": 6.23706579208374, "genuineness": 3.961331605911255, "identity": 0.7172015309333801, "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_NEWS_reg/b17518759e7ba06eca72aace63d85df00156a75085c990eab244efcc7499e903.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": "b18c544bb8abe9ddf58c592d7bba17e860aa94acc6bf1c762cdf09a86ae26e9a", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Natural clear speech.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer we kunnen praten.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c3391af106284107a6b853860b6c669e625913e83c011d6b11ecbdfb32ef2aa9", "generation_seconds_batch": 14.240347187966108, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 6.260893821716309, "genuineness": 4.162164688110352, "identity": 0.583861231803894, "hypothesis": "Je bericht kwam nadat ik vertrok was. Ik las het twee eer keer en legde mijn telefoon weg. Zeg maar wanneer we kunnen praten.", "wer": 0.08695652173913043, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/b18c544bb8abe9ddf58c592d7bba17e860aa94acc6bf1c762cdf09a86ae26e9a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "b1e64e742b4796cf289d5031c37d9442693413068d80c4b363098519555ae168", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "686626219985d48fb4b89f062d8fb945ef8c99c3bfe2034166823ad9077f17c3", "generation_seconds_batch": 20.73820419306867, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7100024223327637, "target_z": 2.2187498343578715, "ce": 6.372091293334961, "genuineness": 3.7918784618377686, "identity": 0.6014213562011719, "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_NEWS_reg/b1e64e742b4796cf289d5031c37d9442693413068d80c4b363098519555ae168.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "b1f58fb1e413f62c1d339a77f9d3b3abecbe9d7a0c1fd2af0651ce73bcc9bf41", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5ade25c51139c606b4f31714b56481c2b560d14c7e30b8c5dea28ed4c3afa185", "generation_seconds_batch": 21.970107111963443, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6497364044189453, "target_z": 2.1093750794406128, "ce": 6.109982967376709, "genuineness": 3.8169827461242676, "identity": 0.45258650183677673, "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_NEWS_reg/b1f58fb1e413f62c1d339a77f9d3b3abecbe9d7a0c1fd2af0651ce73bcc9bf41.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "b30b46cc254507fed950be9188074d14a9b7aa4d06ceaa8c2ce2e6af4082131a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "094cc8b557ef1e870f83dcdb35573cf1719570e8bfaf76240742deac3817d159", "generation_seconds_batch": 18.734575642971322, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8017054796218872, "target_z": 0.5703124738015001, "ce": 6.260893821716309, "genuineness": 3.9550557136535645, "identity": 0.24646112322807312, "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_NEWS_reg/b30b46cc254507fed950be9188074d14a9b7aa4d06ceaa8c2ce2e6af4082131a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "b32f6e91df31434843ef0c3908ccf389ecae3f1e781c702a08e83ed0c9260ec6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ba7a41ea48ee775bc5f207999a1167b5249748761792b457808bbfd0973d058f", "generation_seconds_batch": 18.772262368933298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1891307830810547, "target_z": 1.2734375363398547, "ce": 6.181467056274414, "genuineness": 4.331617832183838, "identity": 0.6142451763153076, "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_NEWS_reg/b32f6e91df31434843ef0c3908ccf389ecae3f1e781c702a08e83ed0c9260ec6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "b3c482466be31c69845143e01cb51ae7a9d929af1522adf98040a91e05739ff4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d54e013af10c991051e9c50a8900df3f9c89eba453db82aaac081d387178d37b", "generation_seconds_batch": 18.972911934019066, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8361432552337646, "target_z": 0.6328124264751777, "ce": 6.23706579208374, "genuineness": 4.369274139404297, "identity": 0.6985530257225037, "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_NEWS_reg/b3c482466be31c69845143e01cb51ae7a9d929af1522adf98040a91e05739ff4.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": "b3f8758355de098819f7c5dcb4123f008d2973acc16b617f98bb9ad1df8077c5", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La route tourne à gauche après le pont.\n(Natural clear speech.) [3.8 seconds duration] Continue... jusqu’à la petite barrière en bois.\n(Natural clear speech.) [3.2 seconds duration] Je t’attendrai de l’autre côté.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "431590543f5a321fb6e353242e670710814fdfd438b9eda967fdd22fcd5483e1", "generation_seconds_batch": 14.118994012009352, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6927835941314697, "target_z": 2.187499966195484, "ce": 6.006728172302246, "genuineness": 4.042920112609863, "identity": 0.8104845285415649, "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_NEWS_reg/b3f8758355de098819f7c5dcb4123f008d2973acc16b617f98bb9ad1df8077c5.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "c902e534de592edd358a451c96b56315df8385dc3b1099308003c02a22c7efdd", "generation_seconds_batch": 15.748168179066852, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8821914196014404, "target_z": 2.531249814075162, "ce": 6.308549880981445, "genuineness": 3.998987913131714, "identity": 0.5698044300079346, "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_NEWS_reg/b429ad869b4f073af736e4e6510807542fc4f5324b0eb7f096c719d9002f4f71.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "b47d9a79b98201b4367aeed4221683b18f9a4b048ed8191630a3a045c5fdf925", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ac3fe30f0f7d0d384a5d47d17ae60d55749c72889d80173772d5667958de062d", "generation_seconds_batch": 21.00167818600312, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6574971675872803, "target_z": 0.30859369210976634, "ce": 6.459460258483887, "genuineness": 4.3504462242126465, "identity": 0.47119006514549255, "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_NEWS_reg/b47d9a79b98201b4367aeed4221683b18f9a4b048ed8191630a3a045c5fdf925.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "b50bfc7a8485eee225ca3ff0a3b4a0ed323dcace9859d058ffe1352ec30f6884", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4a3e8273b834017a694e5c5ee9bc270ee0b89c9b0e2b07e8ee1b33e69bc183af", "generation_seconds_batch": 15.188897399930283, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9997228384017944, "target_z": 0.9296874721112743, "ce": 6.197352409362793, "genuineness": 4.406930446624756, "identity": 0.5924555063247681, "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_NEWS_reg/b50bfc7a8485eee225ca3ff0a3b4a0ed323dcace9859d058ffe1352ec30f6884.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "b53cc7778a6ddd6fb073fa1f1fa8f7ab6e4d2babad3601ec7a5701f4343b50fa", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7ebf0cf54fa8375babf569baf9f412b9fc61834a74fb7a01a864665509f60695", "generation_seconds_batch": 19.61942932603415, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7166872024536133, "target_z": 0.4160156818338425, "ce": 6.332377910614014, "genuineness": 4.557555198669434, "identity": 0.7113292813301086, "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_NEWS_reg/b53cc7778a6ddd6fb073fa1f1fa8f7ab6e4d2babad3601ec7a5701f4343b50fa.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "b58cb6539d28a9fb54f98655d3201b61912b8a53c76380f3c9929964281683d0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cb1c2a0fe0f5aba320f6cce607a31e6a859c467f94e13546c2fa37541f70ed2e", "generation_seconds_batch": 18.764511754969135, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7694200277328491, "target_z": 0.5117187005608953, "ce": 6.23706579208374, "genuineness": 4.042920112609863, "identity": 0.33372005820274353, "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_NEWS_reg/b58cb6539d28a9fb54f98655d3201b61912b8a53c76380f3c9929964281683d0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "b59c0cb9c5f52b159cbca5aba2feb1e5159f706917957f24461ba16d5d873fac", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b6affc22c33fa983c4fe9120422cbc8b70f776d8ed61bfcc973c0237cbaabce9", "generation_seconds_batch": 19.08264987205621, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3139677047729492, "target_z": 1.4999998377383232, "ce": 6.109982967376709, "genuineness": 4.168440818786621, "identity": 0.7053353786468506, "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_NEWS_reg/b59c0cb9c5f52b159cbca5aba2feb1e5159f706917957f24461ba16d5d873fac.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "b59e0ded4625fb85450d202374ab870ee237fd26703c135a2722abb8546b162d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1702, "scale": 0.75, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6104d2059aeb66f8d30e092a7f1e4263245d5c48df5c713d7547b7d3928bf9c8", "generation_seconds_batch": 18.801546472939663, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2063496112823486, "target_z": 1.3046874045022423, "ce": 5.935244083404541, "genuineness": 4.300237655639648, "identity": 0.6877114772796631, "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_NEWS_reg/b59e0ded4625fb85450d202374ab870ee237fd26703c135a2722abb8546b162d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "b5a66b112e369db83e14d38922b7b9b6e935dea2aba39dd3ac7958ff4d3167ba", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ebe29c46aa166cc5334f332e92c05267167ab7d593751c8209bc2b3de000932a", "generation_seconds_batch": 20.533045855001546, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.965285062789917, "target_z": 0.8671875194375968, "ce": 6.41180419921875, "genuineness": 4.030367851257324, "identity": 0.42533066868782043, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/b5a66b112e369db83e14d38922b7b9b6e935dea2aba39dd3ac7958ff4d3167ba.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": "b5e46eae1b2c77afc80fe9467c3fe2e1b377887777edba7843c4060c1a83f673", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "f3cdc74cbd1af6e08ad6bf4f4cef506c84416d5e6c1c47e406cc386438b5f709", "generation_seconds_batch": 14.84269108902663, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4818520545959473, "target_z": 1.8046874585894679, "ce": 6.133810997009277, "genuineness": 4.118232727050781, "identity": 0.7288758754730225, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/b5e46eae1b2c77afc80fe9467c3fe2e1b377887777edba7843c4060c1a83f673.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "b5fdbe00be5cefdc73b785fd71b0766f3055fb6feaf4bc9951f9f6f4005ed6a7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "1177b8f4651c762ae705a473d2f97b17a15c23f8d7fa09dbad9f17326f6b832b", "generation_seconds_batch": 14.307129143038765, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0944268703460693, "target_z": 1.1015626124000157, "ce": 6.173524379730225, "genuineness": 4.369274139404297, "identity": 0.4460514187812805, "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_NEWS_reg/b5fdbe00be5cefdc73b785fd71b0766f3055fb6feaf4bc9951f9f6f4005ed6a7.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": "b60fb7031a92b56df1c2b0c2607731b4b8d47d45ee87ff5c2c5d772977fae1b2", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Straße führt hinter der Brücke nach links.\n(Natural clear speech.) [3.8 seconds duration] Geh weiter... bis zu dem kleinen Holztor.\n(Natural clear speech.) [3.2 seconds duration] Ich warte auf der anderen Seite.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b416229f58a07d123fda710ebd79422fb922f34f92904c4142d76a68a3a98f9e", "generation_seconds_batch": 14.542624247027561, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0113332271575928, "target_z": 2.765624907037581, "ce": 6.3641486167907715, "genuineness": 4.18099308013916, "identity": 0.5116663575172424, "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_NEWS_reg/b60fb7031a92b56df1c2b0c2607731b4b8d47d45ee87ff5c2c5d772977fae1b2.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": "b64805e2de3484004c54196edc9f61d63fd32d3d9020e04134ea10cae37a72ea", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "6762fb35f92a479e5cff322efd2e5dfbc578281a43961fbca0f9675b5ba1d5c0", "generation_seconds_batch": 14.627989643951878, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6239080429077148, "target_z": 2.0625000608481288, "ce": 6.292664527893066, "genuineness": 4.143336772918701, "identity": 0.34902387857437134, "hypothesis": "Les clés sont sur la table de la cuisine. J'y regardais dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.08333333333333333, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/b64805e2de3484004c54196edc9f61d63fd32d3d9020e04134ea10cae37a72ea.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": "b6fb09fd23da787f0d3a979387a708f73f88c88c8ebe75510c39e31646f08702", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "adff7700ac3740ddc685ee5f0a1806393bcfd2a42eb181a703ae3395b91a2c65", "generation_seconds_batch": 15.491098306025378, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.942457675933838, "target_z": 2.6406250016902257, "ce": 6.38797664642334, "genuineness": 4.024092197418213, "identity": 0.6004689335823059, "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_NEWS_reg/b6fb09fd23da787f0d3a979387a708f73f88c88c8ebe75510c39e31646f08702.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "b71ae6c5febcb9e0b612f63c215e48abd3d2827c532515a28d18a743dcce3fb2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "3041ded9fc96a8480baed6565651ce3d49a2567959d89f09714549a050ccedb3", "generation_seconds_batch": 16.103394269011915, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.016941785812378, "target_z": 0.9609375566225643, "ce": 6.09409761428833, "genuineness": 4.432034492492676, "identity": 0.655957818031311, "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_NEWS_reg/b71ae6c5febcb9e0b612f63c215e48abd3d2827c532515a28d18a743dcce3fb2.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": "b792aaf99a4f5d952e582e70ece44696f33ad53255008e0887ec0181a5345645", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "39648f4407a958e6ac3f848531c74a25bbd21201a3ca97fd18a4c20ae5528708", "generation_seconds_batch": 19.44807394396048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4258906841278076, "target_z": 1.7031250625383545, "ce": 6.24500846862793, "genuineness": 4.369274139404297, "identity": 0.5867775678634644, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in Della 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_NEWS_reg/b792aaf99a4f5d952e582e70ece44696f33ad53255008e0887ec0181a5345645.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "b7b69d813a747851a48d43067e566d8902d64c9b3077cf9003a1cf697f78ff4b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 0.75, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0da39f026e00488bc12ebed4c267b16d803ec106d06663d7814878da917ae236", "generation_seconds_batch": 20.372638594009914, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.503375768661499, "target_z": 1.843750118315806, "ce": 6.0305562019348145, "genuineness": 4.15588903427124, "identity": 0.3416557312011719, "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_NEWS_reg/b7b69d813a747851a48d43067e566d8902d64c9b3077cf9003a1cf697f78ff4b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "b7f8a033b7dce77f6d6fd6aff67d01c2543ab980a75e19241914fe2c344c8ec1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "238d2a3404dec73caf2f583279353f41321f8b39735fda86cb32c690f1c4ccc4", "generation_seconds_batch": 19.951832809019834, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5808608531951904, "target_z": 1.9843751740932574, "ce": 5.927301406860352, "genuineness": 4.105680465698242, "identity": 0.6866284608840942, "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_NEWS_reg/b7f8a033b7dce77f6d6fd6aff67d01c2543ab980a75e19241914fe2c344c8ec1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "b82d4da53983e5e585796f693f17691c84ed928ea6252c576c223eff12f60631", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "c50eb24473681aa57344c89261923d7765adc4818e3efd88c6b918ccbcf399d5", "generation_seconds_batch": 21.122083690017462, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8318386077880859, "target_z": 0.625000067609032, "ce": 6.23706579208374, "genuineness": 3.9550557136535645, "identity": 0.2506794035434723, "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_NEWS_reg/b82d4da53983e5e585796f693f17691c84ed928ea6252c576c223eff12f60631.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "b8c8205b2ab90c8eeca5d30427bc6f6473d28fb27e51d56691437b931a39f06e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d09a1e67af9cd7654841d98e35fd453d78fb718285395b0678e47834c09bad29", "generation_seconds_batch": 19.085687605896965, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0199427604675293, "target_z": 2.7812500574676773, "ce": 6.09409761428833, "genuineness": 4.124508857727051, "identity": 0.7247161865234375, "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_NEWS_reg/b8c8205b2ab90c8eeca5d30427bc6f6473d28fb27e51d56691437b931a39f06e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "b94df41f1c304d3b9fafe6ed15879519e043062546ec7aa743213cdb7341dafe", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8d6abec53c04a1e186f459593f986a6a7cb371ef918225cb61d9f44a075e26e1", "generation_seconds_batch": 18.725731721962802, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9308472871780396, "target_z": 0.8046875667639191, "ce": 6.173524379730225, "genuineness": 3.7793264389038086, "identity": 0.371515154838562, "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_NEWS_reg/b94df41f1c304d3b9fafe6ed15879519e043062546ec7aa743213cdb7341dafe.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "b96b567a00f83be7861db596031f4c59b4ce025cf9bec718bf9cb18a2a4f38c5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a2a85f6a612595408f919e527cfbca80ae3e9b9fc82a56070f4a9f6dcba1246a", "generation_seconds_batch": 20.138614128925838, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4043669700622559, "target_z": 1.6640624028120166, "ce": 6.046441555023193, "genuineness": 4.118232727050781, "identity": 0.6569130420684814, "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_NEWS_reg/b96b567a00f83be7861db596031f4c59b4ce025cf9bec718bf9cb18a2a4f38c5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "b9f06c35b82d4aabbe6be58ba987f70ca6070241f84430f95c5c0b90e6374924", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2638d69dbb5e1860c780f56e683874983e8a16509801704ad5d0d0009030e06f", "generation_seconds_batch": 18.957389133051038, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1288645267486572, "target_z": 1.164062348724791, "ce": 6.2767791748046875, "genuineness": 4.406930446624756, "identity": 0.823434054851532, "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_NEWS_reg/b9f06c35b82d4aabbe6be58ba987f70ca6070241f84430f95c5c0b90e6374924.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "ba0b00dc5ab2c23d377ba86f235a353ac48717df0376a654f332983f51298028", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Tell me when you are ready.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The door is still open.\n(Natural clear speech.) [3.8 seconds duration] I put your coat on the chair.\n(Natural clear speech.) [3.2 seconds duration] Tell me when you are ready.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a07835150e7d41880e77b90c6b58cf8a1a306d940295ee1e48b93544385e73c1", "generation_seconds_batch": 15.017885050037876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0556843280792236, "target_z": 1.03125008451129, "ce": 6.24500846862793, "genuineness": 3.9425034523010254, "identity": 0.28301164507865906, "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_NEWS_reg/ba0b00dc5ab2c23d377ba86f235a353ac48717df0376a654f332983f51298028.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "ba31f66b595729e1a218ae0ddd438b00a197f60bc2ce1236390ede1f8c959d35", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4fcbf1f2294593fae90a44b6037568083f24ec7a9f2c7af3d62bcbe5c12d462b", "generation_seconds_batch": 19.08264987205621, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.460328459739685, "target_z": 1.7656250152120323, "ce": 6.197352409362793, "genuineness": 4.394378662109375, "identity": 0.5844236612319946, "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_NEWS_reg/ba31f66b595729e1a218ae0ddd438b00a197f60bc2ce1236390ede1f8c959d35.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": "ba4e37e00be28917de65d3d46f5149c06a88f8c6cb07e177f73dc44fa2fabdc7", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3b2d4fb2011b87fac46bb07cd3d4e8ec0126e0fb5246691ad9599672fcbbd9dc", "generation_seconds_batch": 20.000740652903914, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.67148756980896, "target_z": 0.33398439929699586, "ce": 6.467403411865234, "genuineness": 4.369274139404297, "identity": 0.5062212347984314, "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_NEWS_reg/ba4e37e00be28917de65d3d46f5149c06a88f8c6cb07e177f73dc44fa2fabdc7.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "76383a82bca122a277b6f59157d76ea736cbffac4a531b75c1ff2226e84eb1d2", "generation_seconds_batch": 14.338822501944378, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0556843280792236, "target_z": 1.03125008451129, "ce": 6.181467056274414, "genuineness": 4.545003414154053, "identity": 0.7111878395080566, "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_NEWS_reg/ba70e7e208a6b15a9236ccd5a69e0e5fd403df2a937202d690effa2cdc45d9f2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "bab68cb9aaf2c31b15d7faee7b99390925a89c88eab37f71efcfd1ebf6509b7a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "bb595b5c9e04eee213b6b6bab76f66e4defb88222b0010563da85c4db3821299", "generation_seconds_batch": 18.82938202598598, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0126370191574097, "target_z": 0.9531249814075162, "ce": 6.268836498260498, "genuineness": 4.432034492492676, "identity": 0.6465028524398804, "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_NEWS_reg/bab68cb9aaf2c31b15d7faee7b99390925a89c88eab37f71efcfd1ebf6509b7a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "bad507f46afc24d9722551307c8473c69bae2eff93d77709225ae1dea8df5a7f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "1b0b57dbb22f41639f925023e846064d2ee738874ce2d7a0c8a6fd09ff9e1c07", "generation_seconds_batch": 18.7434428720735, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.607454776763916, "target_z": 0.21777345176128018, "ce": 6.300607204437256, "genuineness": 4.118232727050781, "identity": 0.3884595036506653, "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_NEWS_reg/bad507f46afc24d9722551307c8473c69bae2eff93d77709225ae1dea8df5a7f.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": "bb0193bfd0e1fb93cb197d499d8784701e9f46fccc6be5e0cc7af987edd1b55d", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b0b0b763078c8a385c39d1fed2ec5b347243fda0279520cce9cc2b0df6519a4b", "generation_seconds_batch": 14.22829932696186, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6927835941314697, "target_z": 2.187499966195484, "ce": 6.268836498260498, "genuineness": 4.13078498840332, "identity": 0.5359302759170532, "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_NEWS_reg/bb0193bfd0e1fb93cb197d499d8784701e9f46fccc6be5e0cc7af987edd1b55d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "bb19431e42ff8a54b78b8f6a04e9c47db8f72f9b8099f79ae43a15d106fced69", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8f3c746a78bfb7aa54f132a647fd9d7c1e3fa31d74bf543f90dd6b06911bcb76", "generation_seconds_batch": 19.71350035793148, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6669552326202393, "target_z": 2.1406249476030004, "ce": 6.141753673553467, "genuineness": 3.8295347690582275, "identity": 0.4449734091758728, "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_NEWS_reg/bb19431e42ff8a54b78b8f6a04e9c47db8f72f9b8099f79ae43a15d106fced69.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": "bb2b7438b3cc51ed7a6dfeb177a48fb0f2e55069dd944275dd7783270c7b1475", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "074134033b0a5ac6c69d5eb48b07351fdbace8cf672c0fe6c0ec323add762067", "generation_seconds_batch": 15.518799489014782, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4504151344299316, "target_z": 3.5624997904120006, "ce": 6.109982967376709, "genuineness": 3.810706615447998, "identity": 0.7310801148414612, "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_NEWS_reg/bb2b7438b3cc51ed7a6dfeb177a48fb0f2e55069dd944275dd7783270c7b1475.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "bb335026c4cc36e8216a3847628232507e546a6e25a04ca6f086cee58131b6d3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "da19cb58b078a117343894b70d13ee0a651578fd7cf8f7e62f943a35c1bef647", "generation_seconds_batch": 20.591081165941432, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3311867713928223, "target_z": 1.5312501385985156, "ce": 6.292664527893066, "genuineness": 4.362998008728027, "identity": 0.8096485733985901, "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_NEWS_reg/bb335026c4cc36e8216a3847628232507e546a6e25a04ca6f086cee58131b6d3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "bb99ac36e25deaf993200ed143c1a509db6a43983e90c507a6507813ea2f9bc3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "35782bd0ad2b5ae17035964bf3829b2f00cdb6b4ea75bf67e6797226258d5a0c", "generation_seconds_batch": 19.98474968795199, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0371615886688232, "target_z": 2.812499925630065, "ce": 6.054384231567383, "genuineness": 4.080576419830322, "identity": 0.7408213019371033, "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_NEWS_reg/bb99ac36e25deaf993200ed143c1a509db6a43983e90c507a6507813ea2f9bc3.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": "bba8cb0d631674e5e3aff1d68265b35847b3bf942f7a0dee8a8efccb93564361", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Prends-les quand tu partiras.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Les clés sont sur la table de la cuisine.\n(Natural clear speech.) [3.8 seconds duration] J’ai regardé dans le tiroir... puis près de la fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Prends-les quand tu partiras.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "244de4d4e8ceeb88fd86b9d3f3b04a1673f3811dbcc718289382ee1a5f664f4a", "generation_seconds_batch": 14.627989643951878, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4258906841278076, "target_z": 1.7031250625383545, "ce": 6.252951145172119, "genuineness": 4.268857479095459, "identity": 0.6083414554595947, "hypothesis": "Les clés sont sur la table de la cuisine. J'ai regardé dans le tiroir, puis près de la fenêtre. Prends-les quand tu partiras.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/bba8cb0d631674e5e3aff1d68265b35847b3bf942f7a0dee8a8efccb93564361.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "bbbc57de27d2d36523aa193278cfc55a010afef0b74496db110fba405bc81aea", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "77809cf85cdce0dbd677c3890ffa09c7a22e182cd3e19773e197d8503c237399", "generation_seconds_batch": 18.927428891998716, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6833255290985107, "target_z": 0.35546871070225017, "ce": 6.419747352600098, "genuineness": 4.413206577301025, "identity": 0.3188053071498871, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/bbbc57de27d2d36523aa193278cfc55a010afef0b74496db110fba405bc81aea.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "bbcf3f763c985f687f407ab5b815e58214d325e348765fc0e482f8a5c58e7564", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "858e62cf2250e0af38fe333a830e7f98e3bf46806f5165d9ebdedf7f450e9764", "generation_seconds_batch": 18.782732837018557, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0944268703460693, "target_z": 1.1015626124000157, "ce": 6.41180419921875, "genuineness": 4.406930446624756, "identity": 0.638529896736145, "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_NEWS_reg/bbcf3f763c985f687f407ab5b815e58214d325e348765fc0e482f8a5c58e7564.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "bc44c97daf13c904a13bbd8f0da9184ac9d9b3cff0a76f91da49091d78e521d4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "51a88cfe3b31f2e8c8fed89123258b301c0735f74e3b3d2c314f516a211b190c", "generation_seconds_batch": 15.079204719979316, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3040544986724854, "target_z": 3.2968748292871943, "ce": 6.300607204437256, "genuineness": 4.256305694580078, "identity": 0.45500320196151733, "hypothesis": "Ik heb het kleine doosje gevonden. Het stond naast het eneelsduk DRA, 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_NEWS_reg/bc44c97daf13c904a13bbd8f0da9184ac9d9b3cff0a76f91da49091d78e521d4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "bce38e7fb947e42ba3781f32ab6312618d136640c2473a6b3c7d894f1d0514c1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a64c5b6fb3579610d577ab58919e4f027aa715646ce27e36204ef0456383a5b2", "generation_seconds_batch": 19.650221711955965, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5076804161071777, "target_z": 1.8515624771819517, "ce": 6.006728172302246, "genuineness": 4.18726921081543, "identity": 0.7423639297485352, "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_NEWS_reg/bce38e7fb947e42ba3781f32ab6312618d136640c2473a6b3c7d894f1d0514c1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "bd1a90edac8f9579e1e94b447507fe9077c130ad58b926df5b3d0a451464d91d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fbca9cfebc4659f2bad827bf04cc487016d3c1f23a2c5a4cd57582c5c40e3df0", "generation_seconds_batch": 21.49835668702144, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4775474071502686, "target_z": 1.7968750997233223, "ce": 6.252951145172119, "genuineness": 4.331617832183838, "identity": 0.6632130146026611, "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_NEWS_reg/bd1a90edac8f9579e1e94b447507fe9077c130ad58b926df5b3d0a451464d91d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "bd99369d1bbc00d88846ccca1c56ca40d96c6738ad9661829c0cbe22bb5ffc42", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "afb22edc32a2618e5b03b239ffaff47521f5e60bb4c3a4a37b5ac40363581f21", "generation_seconds_batch": 16.310151770012453, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5507276058197021, "target_z": 1.929687363936823, "ce": 6.149696350097656, "genuineness": 4.457139015197754, "identity": 0.659833550453186, "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_NEWS_reg/bd99369d1bbc00d88846ccca1c56ca40d96c6738ad9661829c0cbe22bb5ffc42.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": "be0aff980c55bf4a17e3c656b4300fe57a1b9153172222fb088c8a9b6627b6ea", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "bdea931d678db794cae7e3729de7ab99577b114b50dd8dd378dec5641ef4b9af", "generation_seconds_batch": 14.289620777009986, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.5192909240722656, "target_z": 3.687500128457161, "ce": 6.316492557525635, "genuineness": 4.093128204345703, "identity": 0.670025110244751, "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_NEWS_reg/be0aff980c55bf4a17e3c656b4300fe57a1b9153172222fb088c8a9b6627b6ea.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": "bf0bd6efd0101661d5a870a02468e0778202b83126254bb19f316ebe1965ed4f", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Someone moved the little blue chair.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "dc609fc1f96d27ff856a147c2e76aceaa9fd6a5c8c6599bae84bf55768a33a98", "generation_seconds_batch": 15.76106071996037, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.483288764953613, "genuineness": 4.293961524963379, "identity": 0.21860617399215698, "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_NEWS_reg/bf0bd6efd0101661d5a870a02468e0778202b83126254bb19f316ebe1965ed4f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "bf12408539a4df7f774c1579c6ca0dc235a7da19b6a4c49db1aae0f7d2251cd4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "968766ea3bb3c76f6f66a4945a0e8db7364a419796d1ab57b0032370ae88d0e0", "generation_seconds_batch": 18.85329126601573, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4215859174728394, "target_z": 1.6953124873233065, "ce": 6.252951145172119, "genuineness": 4.2374773025512695, "identity": 0.6639657616615295, "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_NEWS_reg/bf12408539a4df7f774c1579c6ca0dc235a7da19b6a4c49db1aae0f7d2251cd4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "bf7f25bca47754d3034d9f4ab39d12d6d0a94a6ca046507d94293328f9ac3154", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "57157de3cbde4651305f92553b4913fe8ddf2056be510720353def17888cc8fc", "generation_seconds_batch": 21.057972837006673, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1245598793029785, "target_z": 1.1562499898586451, "ce": 6.133810997009277, "genuineness": 4.306513786315918, "identity": 0.7357339262962341, "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_NEWS_reg/bf7f25bca47754d3034d9f4ab39d12d6d0a94a6ca046507d94293328f9ac3154.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "bff177f91f7884767d6b5391ae4bd274afcabbb7b3cd00fc59c68fcb49b9e6bc", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ab55f157b4be4457a2f5bd0f1b417bd9ce1fd36e357a0fc1909c2a87b98f04bd", "generation_seconds_batch": 20.778280753991567, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7960970401763916, "target_z": 2.375000040565419, "ce": 6.181467056274414, "genuineness": 4.15588903427124, "identity": 0.4289153516292572, "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_NEWS_reg/bff177f91f7884767d6b5391ae4bd274afcabbb7b3cd00fc59c68fcb49b9e6bc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "c03f74f58a362cf37ff4af9618a38482817b1b960c1a92bef5e4883dbf686ea8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "dc5662404eabb76c419fb44087d738b37c80d888a9b46112c682844548e79b1a", "generation_seconds_batch": 15.57495532091707, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4990708827972412, "target_z": 1.8359373267518555, "ce": 6.070269584655762, "genuineness": 4.13078498840332, "identity": 0.7693487405776978, "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_NEWS_reg/c03f74f58a362cf37ff4af9618a38482817b1b960c1a92bef5e4883dbf686ea8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "c077578cfa36dd457b41d88ec44d10765a642a6733234d548a63154b47bec5be", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "08e3ea6705903008395dbc77ab821a46c5c1d48dc1bcbf1bbc8616061013754e", "generation_seconds_batch": 16.003847575979307, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8684287071228027, "target_z": 0.6914061997157824, "ce": 6.221180438995361, "genuineness": 4.545003414154053, "identity": 0.8155927658081055, "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_NEWS_reg/c077578cfa36dd457b41d88ec44d10765a642a6733234d548a63154b47bec5be.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "c0b0c8640cb9e7182670c3fe7105fe6d767c4ed9e783b0dd70e42ae63592d99b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b9693985e6600b2f1e85e094cbfa98e5f1051f6893a1dd1db4bfeee9d37b4621", "generation_seconds_batch": 15.65497026196681, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0556843280792236, "target_z": 1.03125008451129, "ce": 6.1894097328186035, "genuineness": 4.519899368286133, "identity": 0.369501531124115, "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_NEWS_reg/c0b0c8640cb9e7182670c3fe7105fe6d767c4ed9e783b0dd70e42ae63592d99b.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": "c0b46fda3f1c959a637412dcb1ef659ada7038a3c6238edb6cd3476ed04cf40c", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "22c87d3a1053a9a45f53666f5830c7fdfd8e7d6941e28acc7d1b280a62f4a42b", "generation_seconds_batch": 14.36836458207108, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.460328459739685, "target_z": 1.7656250152120323, "ce": 5.959072113037109, "genuineness": 4.055472373962402, "identity": 0.7446898818016052, "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_NEWS_reg/c0b46fda3f1c959a637412dcb1ef659ada7038a3c6238edb6cd3476ed04cf40c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "c11f44a8c35afaa27f9affe780fbce4dca473db142edb17ab8926354f7dc865d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9943a8445f4ae10c03e85fed9a4c90730fb421841df5f42c86a04922407cce71", "generation_seconds_batch": 18.731700581032783, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.468937873840332, "target_z": 1.7812499492932259, "ce": 6.117925643920898, "genuineness": 4.105680465698242, "identity": 0.23308467864990234, "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_NEWS_reg/c11f44a8c35afaa27f9affe780fbce4dca473db142edb17ab8926354f7dc865d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "c16a0e049d30815e024782640236dcba43ce5825c234af2a7779e6f6b80be809", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9468c89f9527a5c89296763d804395639abe852968493ebcfd138317c15cbab7", "generation_seconds_batch": 19.023266934091225, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6669552326202393, "target_z": 2.1406249476030004, "ce": 5.911416053771973, "genuineness": 4.325342178344727, "identity": 0.6612023115158081, "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_NEWS_reg/c16a0e049d30815e024782640236dcba43ce5825c234af2a7779e6f6b80be809.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "c1a317b19f3eb184e9043d59d95aff1a46a059b0cd43cc6329155bdbfb451188", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9590fa49c901f2fe04b8e8294f4d1c4d61a0da8aca5585b27b6582aae398bad2", "generation_seconds_batch": 18.735959362005815, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2623111009597778, "target_z": 1.406250016902258, "ce": 6.23706579208374, "genuineness": 4.306513786315918, "identity": 0.7111660838127136, "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_NEWS_reg/c1a317b19f3eb184e9043d59d95aff1a46a059b0cd43cc6329155bdbfb451188.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "c252098ca2472f19f73176ebc5018528df7a0acab06cb4b41b2bfacd0ab8d9c8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fddc42a891e6c88575174377811c83bd218cd29a9b7bc14a8e56132ff690d4c9", "generation_seconds_batch": 18.80143793602474, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5292041301727295, "target_z": 1.8906251369082898, "ce": 6.0146708488464355, "genuineness": 4.118232727050781, "identity": 0.7700093984603882, "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_NEWS_reg/c252098ca2472f19f73176ebc5018528df7a0acab06cb4b41b2bfacd0ab8d9c8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "c296a5b2b84a796f34e19db6b81e0bb22742e255cb0cf6c032979f712e5d2715", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f7f32591615153afa03979a6d102c886b5a4b8073028fbc505be0cf31204f04f", "generation_seconds_batch": 18.8837194229709, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.403861999511719, "genuineness": 4.118232727050781, "identity": 0.5509883165359497, "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_NEWS_reg/c296a5b2b84a796f34e19db6b81e0bb22742e255cb0cf6c032979f712e5d2715.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "c2c391e4b4710cacfb5a48a5ac63f31668c6e0faf29c5b69732e204e23b8db21", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4bf0a12b51160e0746999fee8126d68c9a135dfff36e3d00c56523a9f3c7eb53", "generation_seconds_batch": 19.983958607073873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.252951145172119, "genuineness": 3.8671910762786865, "identity": 0.37954607605934143, "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_NEWS_reg/c2c391e4b4710cacfb5a48a5ac63f31668c6e0faf29c5b69732e204e23b8db21.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": "c2e0a0ddffcaa7c8e443825f99f65a1a93d9f1327acdda3ca8d7c7d3ee43b2cb", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Quelqu’un a déplacé la petite chaise bleue.\n(Natural clear speech.) [3.8 seconds duration] Elle était ici... près de la grande bibliothèque.\n(Natural clear speech.) [3.2 seconds duration] Nous la remettrons après le déjeuner.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "00530c6207cba0eb9001d2b1b0cd69e6da1537324e4a522a8fbef74a08aa9866", "generation_seconds_batch": 13.920932088978589, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.157693862915039, "target_z": 3.0312498681623876, "ce": 6.078212261199951, "genuineness": 4.124508857727051, "identity": 0.7474246025085449, "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_NEWS_reg/c2e0a0ddffcaa7c8e443825f99f65a1a93d9f1327acdda3ca8d7c7d3ee43b2cb.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "c2fe5901bfed367c8e3bae24f8f40dff6592fb7f71756daa0983b34266cc3db6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c9d22d23286f85d16ef414c77128ea36c11878633c2f51b33f0974f61e8c71cf", "generation_seconds_batch": 18.88016403396614, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9674373865127563, "target_z": 0.8710936988706696, "ce": 6.197352409362793, "genuineness": 4.457139015197754, "identity": 0.6924775838851929, "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_NEWS_reg/c2fe5901bfed367c8e3bae24f8f40dff6592fb7f71756daa0983b34266cc3db6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "c303a2443c86b8a526184f05d51dcfd5a0dd31010c68614d56f4c4df7d17e2fc", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3ea1fe10f50f6def896bc9a6752375628c4184d6ba470feb2e46ce96ff94291b", "generation_seconds_batch": 20.51661712594796, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5765559673309326, "target_z": 1.9765623825293068, "ce": 6.308549880981445, "genuineness": 4.356721878051758, "identity": 0.43738988041877747, "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_NEWS_reg/c303a2443c86b8a526184f05d51dcfd5a0dd31010c68614d56f4c4df7d17e2fc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "c30651d443c81c734152859c7b3d3ec2203e2bfbc139395ec4503f92948b441d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "84a0d8f40bfab28c8dcffc07c815b9419c4d9501fd749aaef54677e50a921fa7", "generation_seconds_batch": 19.61499005497899, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0944268703460693, "target_z": 1.1015626124000157, "ce": 6.229123115539551, "genuineness": 4.381826400756836, "identity": 0.5864137411117554, "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_NEWS_reg/c30651d443c81c734152859c7b3d3ec2203e2bfbc139395ec4503f92948b441d.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": "c386921e4d9017c01b212444336a15309d25124ce8dc469f3e581879852ff779", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Someone moved the little blue chair.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can put it back after lunch.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Someone moved the little blue chair.\n(Natural clear speech.) [3.8 seconds duration] It used to stand here... beside the tall bookshelf.\n(Natural clear speech.) [3.2 seconds duration] We can put it back after lunch.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a34dcca0ad23ad58fec32d9bdf9ecdc6df5dcbcbc5239bd7fd1c750f3c79c37f", "generation_seconds_batch": 14.26039441791363, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7960970401763916, "target_z": 2.375000040565419, "ce": 6.38003396987915, "genuineness": 4.20609712600708, "identity": 0.2764264643192291, "hypothesis": "Someone moved the little blue chair. It used to stand here, beside the tall bookshelf. We can put it back after lunch.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/c386921e4d9017c01b212444336a15309d25124ce8dc469f3e581879852ff779.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "e7cca09b20c9fcb4ab650dcf53eb8a608d6ca526d1e0b2d55a6cb07b1175db64", "generation_seconds_batch": 15.472013409947976, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2881394624710083, "target_z": 1.4531250354947418, "ce": 5.982900142669678, "genuineness": 4.331617832183838, "identity": 0.7328565120697021, "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_NEWS_reg/c3977db0634952a7f9e4083e6135fbbc87199d34c7cb8c3338515d9ae38c18f4.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": "c43b3b7d24b0201a56debac3f4e828769592b1ab7055b5951d636b1a6a416e80", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cb083db7195325165e66d2a33ea664250c8a077b08190d11d3aed0c26b90ad98", "generation_seconds_batch": 16.028420988004655, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2364827394485474, "target_z": 1.3593749983097743, "ce": 6.443574905395508, "genuineness": 4.494794845581055, "identity": 0.6010942459106445, "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_NEWS_reg/c43b3b7d24b0201a56debac3f4e828769592b1ab7055b5951d636b1a6a416e80.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "e636a8288201aed4e73e29d04c5d18b98fecced7f309922cdc5740fdec1113c1", "generation_seconds_batch": 15.062218261999078, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.468937873840332, "target_z": 1.7812499492932259, "ce": 6.221180438995361, "genuineness": 4.20609712600708, "identity": 0.7496234178543091, "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_NEWS_reg/c493619abd8899c9b3ce96f2a00a6bcdd900e874ef21436707b986365a8437ee.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "c4a08f79a0d77e95d28f8e2f77bbd750209a51516a1bd4038f9e95244a3b6885", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5869b770a392628eec68c1adffb3616c796aa6db56e429d071adc0a41be130b1", "generation_seconds_batch": 19.600245298002847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6768684387207031, "target_z": 0.3437499560541292, "ce": 6.443574905395508, "genuineness": 4.494794845581055, "identity": 0.34128546714782715, "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_NEWS_reg/c4a08f79a0d77e95d28f8e2f77bbd750209a51516a1bd4038f9e95244a3b6885.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "c4d18912e9b7866284e4d69eaa0b15846e8aa54a92c825d5316db5368a89d40a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a63932a2732aa2eec60b3afb64d23ec0198349ed5d8ab06e62a05723e8fa9b29", "generation_seconds_batch": 21.20744886796456, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.403518408536911, "target_z": -0.15234375633834676, "ce": 6.316492557525635, "genuineness": 4.507347106933594, "identity": 0.11327657103538513, "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_NEWS_reg/c4d18912e9b7866284e4d69eaa0b15846e8aa54a92c825d5316db5368a89d40a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "c58805ca85e34da05c3a6fc7d425d2cced5184a8b1122d57fedc0af51334368e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "df14278b7f2df790a7df4e65d61e461ceb26926967a0cc155c93386466c5796f", "generation_seconds_batch": 20.610676417010836, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.573555052280426, "target_z": 0.1562499898586452, "ce": 6.165581703186035, "genuineness": 4.193545341491699, "identity": 0.20010453462600708, "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_NEWS_reg/c58805ca85e34da05c3a6fc7d425d2cced5184a8b1122d57fedc0af51334368e.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": "c5fb88073669ad6d76e8bd3c5010964d8cb33b2b839b84bab25f00bafa32b7e6", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo podemos hablar.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Tu mensaje llegó después de que me fuera.\n(Natural clear speech.) [3.8 seconds duration] Lo leí dos veces... y guardé el teléfono.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo podemos hablar.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "70a32fe237b549ff50811d109e4055c75fe374b09529d3d600f0ab312894a56f", "generation_seconds_batch": 15.074431804940104, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6669552326202393, "target_z": 2.1406249476030004, "ce": 6.125868320465088, "genuineness": 4.111956596374512, "identity": 0.5074737668037415, "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_NEWS_reg/c5fb88073669ad6d76e8bd3c5010964d8cb33b2b839b84bab25f00bafa32b7e6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "c78cad57882a7212370b972b45c9c7edc51d0d55827bc6ed9fbd11c11c0b922c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7ffcda0f9db1a9e86d957b603807a022eb00db705a44947e137dbc3a9e2ca058", "generation_seconds_batch": 18.707597031025216, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.387148141860962, "target_z": 1.632812534649629, "ce": 6.300607204437256, "genuineness": 4.281409740447998, "identity": 0.7117911577224731, "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_NEWS_reg/c78cad57882a7212370b972b45c9c7edc51d0d55827bc6ed9fbd11c11c0b922c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "c7ad3ffecdbd9c157de98ca9ebe6a42322a05753cb359aceb3eb37c8fc2692c2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9092d8546779ee96f75d4dab81ce8490f16663fd7735bb61da3a08f63c727bfb", "generation_seconds_batch": 19.71063242899254, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9609802961349487, "target_z": 0.8593749442225486, "ce": 6.181467056274414, "genuineness": 3.7667744159698486, "identity": 0.3533020317554474, "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_NEWS_reg/c7ad3ffecdbd9c157de98ca9ebe6a42322a05753cb359aceb3eb37c8fc2692c2.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "f19f197165566511f9ab512fa7a039eba882fae1b438a4b977641fb61e2e5f6e", "generation_seconds_batch": 13.968352608033456, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6833255290985107, "target_z": 0.35546871070225017, "ce": 6.149696350097656, "genuineness": 4.419482707977295, "identity": 0.792515218257904, "hypothesis": "Je bericht kwam nadat ik vertrokken was, ik las het twee keer en legde mijn telefoon weg. Zeg maar wanneer we wat kunnen praten.", "wer": 0.043478260869565216, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/c82131bfaf89850d394bf3a21117deb94436b65306054c3e1cbac0c65e2f913d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "c834d709d879e9e03cd0623207ebbd9827a7ec2214b1c414a0ff5fbc2f2808e1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "76d45147cc9589d97e57b00c2a7df10729c5ecb01b8339d8e0b117c7f64403db", "generation_seconds_batch": 20.607717695995234, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.387148141860962, "target_z": 1.632812534649629, "ce": 6.229123115539551, "genuineness": 4.193545341491699, "identity": 0.355729341506958, "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_NEWS_reg/c834d709d879e9e03cd0623207ebbd9827a7ec2214b1c414a0ff5fbc2f2808e1.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "3e00960061f2678dcddb202dc6c334b08d6ae6111392d54136effce5cf178e3b", "generation_seconds_batch": 14.339589474955574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9846563339233398, "target_z": 0.9023437833819595, "ce": 6.395919322967529, "genuineness": 4.369274139404297, "identity": 0.4106927514076233, "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_NEWS_reg/c8907269f81c0a33f09ba3b9d9f76b6acb2a8250e6881dbc11786cea01719e9d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "c8a71f87187b184524769f73815d9710d1f4918154b349a724bc1ab1dc5a61a4", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "85842f40d726b966a763f90d4f31d6815b731b9dba97f593d6c18f4cc1c49e01", "generation_seconds_batch": 15.422848433954641, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152985095977783, "target_z": 2.0468749104180324, "ce": 5.903473377227783, "genuineness": 3.8922953605651855, "identity": 0.5728884935379028, "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_NEWS_reg/c8a71f87187b184524769f73815d9710d1f4918154b349a724bc1ab1dc5a61a4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "c8ae3625f0ff6a53c565d2e4b37f57af8f40d72f8eb68e9707c4754323e8dbf6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "e262b550c1381bd079e2775b7657e25f495fa6abc7a7c08cd8ce35da889ee853", "generation_seconds_batch": 18.934860236011446, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.090122103691101, "target_z": 1.0937500371849675, "ce": 6.419747352600098, "genuineness": 4.218649387359619, "identity": 0.17371872067451477, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/c8ae3625f0ff6a53c565d2e4b37f57af8f40d72f8eb68e9707c4754323e8dbf6.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": "c9ad3c0acb9bc12d4217bb619175b4e4dd3173740d6dc03b3f26e77d90fa7b17", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai écrit le numéro sur l’enveloppe.\n(Natural clear speech.) [3.8 seconds duration] Lis-le bien... et vérifie les deux derniers chiffres.\n(Natural clear speech.) [3.2 seconds duration] Appelle-moi quand tu auras un moment.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d5ae64dae9db8f948de43f7450e2094ae0f1375976fe4d7bc6af7c4d612d71b9", "generation_seconds_batch": 13.95950996899046, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.1318655014038086, "target_z": 2.9843748495699036, "ce": 6.173524379730225, "genuineness": 4.21237325668335, "identity": 0.5801376700401306, "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_NEWS_reg/c9ad3c0acb9bc12d4217bb619175b4e4dd3173740d6dc03b3f26e77d90fa7b17.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "c9ec6d7ba36c1d6f6ea7423a7c13bef40c2b38b57e5ce5e98254adba33c7c9d2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6ced268f3f1b7976d4bead26791537033deb9c666f9cb452a1baf17b3ca0cae3", "generation_seconds_batch": 21.038800382986665, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3053584098815918, "target_z": 1.4843751200060318, "ce": 6.2767791748046875, "genuineness": 3.8546390533447266, "identity": 0.07366110384464264, "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_NEWS_reg/c9ec6d7ba36c1d6f6ea7423a7c13bef40c2b38b57e5ce5e98254adba33c7c9d2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "c9f1c5bddab5222ef7ce078a09a9c992021de0605bc856af746d6942ea675a34", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6f5f7e44c4ab5b17486c7436fa0aa8de280227c84c47fff8bee1c186bc7ad8f1", "generation_seconds_batch": 18.772262368933298, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5808608531951904, "target_z": 1.9843751740932574, "ce": 5.919358730316162, "genuineness": 3.8546390533447266, "identity": 0.3306480050086975, "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_NEWS_reg/c9f1c5bddab5222ef7ce078a09a9c992021de0605bc856af746d6942ea675a34.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "ca098cc5c1263a62b379be5c9b7cd64082e0664615f19ef61e03eb82fbe2187d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "62c7c8bf87246793d568c67390b8976eea7ec2de54fa9155c8a2ec10545a8600", "generation_seconds_batch": 18.708634820999578, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6927835941314697, "target_z": 2.187499966195484, "ce": 6.419747352600098, "genuineness": 4.055472373962402, "identity": 0.5743505954742432, "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_NEWS_reg/ca098cc5c1263a62b379be5c9b7cd64082e0664615f19ef61e03eb82fbe2187d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "ca1f7a4460b2591d018e155b42d7f9cae6ee57ff134113d3ee40b3ab666b3c56", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fd7ab06fcdf25456eac6912d5c0c69bd8758b1afd8cba18bea7b1bfed53f48ad", "generation_seconds_batch": 14.79409179405775, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5292041301727295, "target_z": 1.8906251369082898, "ce": 6.221180438995361, "genuineness": 4.306513786315918, "identity": 0.5635794401168823, "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_NEWS_reg/ca1f7a4460b2591d018e155b42d7f9cae6ee57ff134113d3ee40b3ab666b3c56.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "ca7cbf8368f0eccc6516f2a9d681b85489cfce59a235b48e12e88871a67299d9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Natural clear speech.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Natural clear speech.) [3.2 seconds duration] Nous pouvons regarder ensemble ce qu’elle contient.", "seed": 1701, "scale": 0.75, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "a27ddb9c32767d36c5e00aec90bfd710c4563e60fe691f0eab50d86c08e18de7", "generation_seconds_batch": 19.293313191039488, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4474142789840698, "target_z": 1.7421875059157903, "ce": 6.165581703186035, "genuineness": 4.15588903427124, "identity": 0.650148868560791, "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_NEWS_reg/ca7cbf8368f0eccc6516f2a9d681b85489cfce59a235b48e12e88871a67299d9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "cae4d50764f714bc0e2c94541c8276bc8b2780914d6f16c3ec056a0e0f1b945b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2ef83e422381cb731f24cd70d87d44bd3f8fba46e8c4aee5868451bd26929208", "generation_seconds_batch": 20.237919704057276, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6841740608215332, "target_z": 2.171874815765388, "ce": 6.133810997009277, "genuineness": 4.093128204345703, "identity": 0.741601824760437, "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_NEWS_reg/cae4d50764f714bc0e2c94541c8276bc8b2780914d6f16c3ec056a0e0f1b945b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "cb0c9f069c4a59211ccaf8a6c0ed93b273372638653d3afc35153e63ca422451", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] He encontrado la cajita.\n(Natural clear speech.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Natural clear speech.) [3.2 seconds duration] Podemos mirar juntos lo que hay dentro.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "516609c89638371edcdcb50214aa4341021ee37088c14e48afcb1ee86086f885", "generation_seconds_batch": 18.844022584031336, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.606689214706421, "target_z": 2.031250192685741, "ce": 6.09409761428833, "genuineness": 4.281409740447998, "identity": 0.5544415712356567, "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_NEWS_reg/cb0c9f069c4a59211ccaf8a6c0ed93b273372638653d3afc35153e63ca422451.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "cb835fbee6ed4db5b3b06b698ccb55b1ed3d9a44e3ea3f0c1790274baa8df69a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "4b13259a64f4dcd20f93ac6875f736e1352d839febd7a6efa43d0910ea539c4c", "generation_seconds_batch": 14.904808830004185, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8490574359893799, "target_z": 0.6562499357714195, "ce": 6.38003396987915, "genuineness": 4.193545341491699, "identity": 0.34992823004722595, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/cb835fbee6ed4db5b3b06b698ccb55b1ed3d9a44e3ea3f0c1790274baa8df69a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "cb84a1a1b547d7ba3bcc0ec4cd6782acf5d7335984d4a5a7d7aee1b2d941a758", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We kunnen er samen in kijken.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Natural clear speech.) [3.8 seconds duration] Het stond naast het oude raam.\n(Natural clear speech.) [3.2 seconds duration] We kunnen er samen in kijken.", "seed": 1701, "scale": 0.75, "cfg": 1.5, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "5a8d3bba3c41cec3cd7cbaf3b033fc788bd18704df8c99b359a6ff26c9e4c921", "generation_seconds_batch": 18.812548773013987, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.821925401687622, "target_z": 2.421875059157903, "ce": 6.24500846862793, "genuineness": 4.080576419830322, "identity": 0.6926898956298828, "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_NEWS_reg/cb84a1a1b547d7ba3bcc0ec4cd6782acf5d7335984d4a5a7d7aee1b2d941a758.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "cb8a8a68a79dae213160facbd3c081b39a1afec5ebf1edf8374e9f2a09314344", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dime cuándo estás listo.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La puerta sigue abierta.\n(Natural clear speech.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Natural clear speech.) [3.2 seconds duration] Dime cuándo estás listo.", "seed": 1701, "scale": 0.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "52859c36198ae5ead8bb7dda5044d3ec9a17a1c689436a45e89ae8db42292104", "generation_seconds_batch": 14.389329454046674, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7444403171539307, "target_z": 2.2812500033804515, "ce": 6.054384231567383, "genuineness": 4.061748504638672, "identity": 0.5837063789367676, "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_NEWS_reg/cb8a8a68a79dae213160facbd3c081b39a1afec5ebf1edf8374e9f2a09314344.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "cb8e0cf25087fec5eb236faf4967e8e53ed46785c763edd0f9efc4785395de4d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1701, "scale": 0.5, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "e1ea941cceeb6dfcd4c8ece855860f18e3537f56d0a2d360f7390598020a4f8b", "generation_seconds_batch": 18.927428891998716, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7844865918159485, "target_z": 0.5390624974646613, "ce": 6.3641486167907715, "genuineness": 4.325342178344727, "identity": 0.32706308364868164, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/cb8e0cf25087fec5eb236faf4967e8e53ed46785c763edd0f9efc4785395de4d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "cb8e1f660704fe51e18cb7c5d86c15b13304470de84143f30b1a8b185bf6d7a2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2b44199617ea2f41b70acd0fcf6aa664985b84ba92436ff3a53530c753516c42", "generation_seconds_batch": 18.811133375042118, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5746312141418457, "target_z": 0.15820307957518162, "ce": 6.141753673553467, "genuineness": 4.758388519287109, "identity": 0.72785884141922, "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_NEWS_reg/cb8e1f660704fe51e18cb7c5d86c15b13304470de84143f30b1a8b185bf6d7a2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "cbeec2be7a94ddca321b4df55d6daf4159f58eefe339514e8ae393d80eeb5453", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "7ed7bb726bbf58ee77211172b3b459eb13cf5ad1ab68c5ce83186df38489897a", "generation_seconds_batch": 18.56931138294749, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6284403204917908, "target_z": 0.2558594043676733, "ce": 6.459460258483887, "genuineness": 4.432034492492676, "identity": 0.38466671109199524, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/cbeec2be7a94ddca321b4df55d6daf4159f58eefe339514e8ae393d80eeb5453.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "cc592f2c4227c8592e57253f9a90587d657a77933cc8d83e150345bc03e6a3f0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f7174656c93a53201f53a42afba667de07146696b4bc441d6d4473670b9a6b9c", "generation_seconds_batch": 20.18418435100466, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.821925401687622, "target_z": 2.421875059157903, "ce": 6.205295085906982, "genuineness": 4.20609712600708, "identity": 0.7325942516326904, "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_NEWS_reg/cc592f2c4227c8592e57253f9a90587d657a77933cc8d83e150345bc03e6a3f0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "cc5aa96043ba3d37d964c855f8ebde96026e69ee736f586a3f6a489a379757bb", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] We can look inside together.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I found the little box.\n(Natural clear speech.) [3.8 seconds duration] It was beside the old window.\n(Natural clear speech.) [3.2 seconds duration] We can look inside together.", "seed": 1702, "scale": 0.0, "cfg": 3.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "73e13074382fc69007b20d64a97e4548073ba6235f593c95671eb9f011476e34", "generation_seconds_batch": 14.620471233967692, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.851209819316864, "target_z": 0.6601562233789436, "ce": 6.41180419921875, "genuineness": 4.356721878051758, "identity": 0.37420418858528137, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/cc5aa96043ba3d37d964c855f8ebde96026e69ee736f586a3f6a489a379757bb.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "cd49e0ebc753c1b2322f8fb8fb77f8f2b85b0beef5d8becb0ff4af4ea466893e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2a3009ba645cecd1be4185c16b8b5f1a3711c54064b43f350e4eef5f8523e00c", "generation_seconds_batch": 15.100355893955566, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7909436821937561, "target_z": 0.5507812521127823, "ce": 6.109982967376709, "genuineness": 4.482243061065674, "identity": 0.6738396286964417, "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_NEWS_reg/cd49e0ebc753c1b2322f8fb8fb77f8f2b85b0beef5d8becb0ff4af4ea466893e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "cd7164a815536c4884b512b26c967a9993cf58f19bf32437ae183f854afe40b7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "68c8873909dcc40721bb3d3f4fd47ae428ac2b8399bc4b9fd0431408b31282ac", "generation_seconds_batch": 19.679421687033027, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 6.149696350097656, "genuineness": 4.231201171875, "identity": 0.6923052072525024, "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_NEWS_reg/cd7164a815536c4884b512b26c967a9993cf58f19bf32437ae183f854afe40b7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "ce405522401bd62ec596f05abb094652064d1daf661e2e2002568669f61bc291", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "933e57d134da1fb89317cd1bcbad4a76b5c440e455d18b3146634e41e37e69be", "generation_seconds_batch": 18.88741986604873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8921047449111938, "target_z": 0.7343750388751934, "ce": 6.221180438995361, "genuineness": 4.3504462242126465, "identity": 0.8232446908950806, "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_NEWS_reg/ce405522401bd62ec596f05abb094652064d1daf661e2e2002568669f61bc291.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "ce6d7488bf0549236d7f1274863769907d65de9fbd8b2a499abf173da51bb8d5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5e3af9503c79e2d45fd53d755609f919da706ffedd83570e0f7f9626d9c6d49e", "generation_seconds_batch": 18.957389133051038, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9222378134727478, "target_z": 0.7890625245082741, "ce": 6.038498878479004, "genuineness": 4.557555198669434, "identity": 0.5077073574066162, "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_NEWS_reg/ce6d7488bf0549236d7f1274863769907d65de9fbd8b2a499abf173da51bb8d5.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "7c9127362392c4b3ed5fb4f694dd1ece57f512522a2fb75a5dc53e1863423c34", "generation_seconds_batch": 14.915260025998577, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 6.38797664642334, "genuineness": 4.457139015197754, "identity": 0.696366012096405, "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_NEWS_reg/cf68e6506caaaad7e7371e72de9d5417fef89211300487ff83acd2484eaab510.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": "cf8c94468a124500ca142932ded2cb1e5bedf1e32ccd679ff517a59b84df0d67", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The road turns left behind the bridge.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Keep going... until you reach the small wooden gate.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "59230978e20af0d2262880b212607b4b5bea7c8eef94617576fb103bd9961480", "generation_seconds_batch": 21.403163607930765, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 6.38003396987915, "genuineness": 3.8922953605651855, "identity": 0.18564674258232117, "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_NEWS_reg/cf8c94468a124500ca142932ded2cb1e5bedf1e32ccd679ff517a59b84df0d67.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "cff89687df3e2a19c95f2b9f0630df3d7aee8635a8d2da477ee4b047637b7013", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e2ef40a1c7048b16ffc43ac1e5f24068180c30a0e3035564451fc609c9bbe4a9", "generation_seconds_batch": 19.831851085997187, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6669552326202393, "target_z": 2.1406249476030004, "ce": 5.982900142669678, "genuineness": 3.7291181087493896, "identity": 0.7424222230911255, "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_NEWS_reg/cff89687df3e2a19c95f2b9f0630df3d7aee8635a8d2da477ee4b047637b7013.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "d040d35cfb7fd7017398f69fa93dfbc4e0eb373fe6fcf6e51e5e39cd246fad47", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "29881995fcf03e7455b5dfd94ad3809335a2de2b4186b7607b8a7751dd986a49", "generation_seconds_batch": 20.816095154033974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.2093505859375, "target_z": 3.124999905347355, "ce": 6.09409761428833, "genuineness": 3.798154592514038, "identity": 0.7309428453445435, "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_NEWS_reg/d040d35cfb7fd7017398f69fa93dfbc4e0eb373fe6fcf6e51e5e39cd246fad47.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "d0ab3d77f3ac546c58cbdcb4886bcaedb7c77df10af0132f4eb6e31a848391d2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "02fde1fbe0459aabc8f0709c739ee9a54966f91ff83b50018248e0f5af0586ec", "generation_seconds_batch": 19.184261474991217, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0858173370361328, "target_z": 1.0859374619699196, "ce": 6.125868320465088, "genuineness": 4.293961524963379, "identity": 0.619570791721344, "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_NEWS_reg/d0ab3d77f3ac546c58cbdcb4886bcaedb7c77df10af0132f4eb6e31a848391d2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "d0da3f1fdeb6f3f81600feeda713762816ec31e7c29405b03a5156cbd51149c6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "d22121ec3a738a0ab3b7e42d29d1b51bee56c849fd207dddb83ebe0f91000877", "generation_seconds_batch": 20.591081165941432, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8382956981658936, "target_z": 0.6367188222571529, "ce": 6.268836498260498, "genuineness": 4.6956281661987305, "identity": 0.7762860059738159, "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_NEWS_reg/d0da3f1fdeb6f3f81600feeda713762816ec31e7c29405b03a5156cbd51149c6.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "3828d4e109470eca4c9042d1c743909809cec6ad1606ebcabebb38b126007d60", "generation_seconds_batch": 14.802104649948888, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1073410511016846, "target_z": 1.1250001216962575, "ce": 6.38003396987915, "genuineness": 4.168440818786621, "identity": 0.30221259593963623, "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_NEWS_reg/d11caac31fa3732e2517803823f1b039213fb31e9438619fd83d8b232c894b7d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "d135fd4fdf1e9e64c465f7ff54fa46fc58545a159de961430cd2cc9d5e036554", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "03380ff34ec0c0eb832c857c6584941a779b87e7c462b095f9700d8bee8b32a6", "generation_seconds_batch": 20.502573780948296, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3311867713928223, "target_z": 1.5312501385985156, "ce": 6.117925643920898, "genuineness": 4.293961524963379, "identity": 0.5084586143493652, "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_NEWS_reg/d135fd4fdf1e9e64c465f7ff54fa46fc58545a159de961430cd2cc9d5e036554.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": "d187db87e573c794786d4d252804f429b66ed7b2234dd96286e502155c69d894", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "7ae6eaf9755967916fd72c9094c0ce2f0ecc5589ba89015e5bf4b248cb98d2b7", "generation_seconds_batch": 13.844444451970048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3311867713928223, "target_z": 1.5312501385985156, "ce": 6.316492557525635, "genuineness": 4.293961524963379, "identity": 0.7220370769500732, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/d187db87e573c794786d4d252804f429b66ed7b2234dd96286e502155c69d894.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "d1e0d81576bd797bd8450849c9f3ee457eca205210ea7ca79057ffecff989f13", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "cb738ee9a550a6b2c8e06dcf79365413241108c24862edbe53b339adb2a579af", "generation_seconds_batch": 18.734744345070794, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6927835941314697, "target_z": 2.187499966195484, "ce": 6.046441555023193, "genuineness": 4.005263805389404, "identity": 0.5486156940460205, "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_NEWS_reg/d1e0d81576bd797bd8450849c9f3ee457eca205210ea7ca79057ffecff989f13.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "d21899011f446f8754373735da54c138a8ec5a8d11abeb84972e363ae6f6f694", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c3540aeebe603c6ad8c0b86ce873d9331746483b9140955554b11686829cd42a", "generation_seconds_batch": 19.963497626013122, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9179331064224243, "target_z": 0.7812500574676772, "ce": 6.268836498260498, "genuineness": 3.9299514293670654, "identity": 0.3162795305252075, "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_NEWS_reg/d21899011f446f8754373735da54c138a8ec5a8d11abeb84972e363ae6f6f694.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "15eb11c7c65ad66f6072d5223a9589d68b5d9070d8f0be13269ebdf49ae6a2fc", "generation_seconds_batch": 15.458260831888765, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8124673366546631, "target_z": 0.5898438036646692, "ce": 6.205295085906982, "genuineness": 4.65797233581543, "identity": 0.7116348743438721, "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_NEWS_reg/d2319f0ba4d4a13f099c61fd2524d2cd47bb09bca092815ed8a29e7f5b0c527c.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": "d2e1e854a388d1e1db1cf96c3328ba19d06bd3d89bdf749d1d1ae643c07c8268", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I wrote the number on the envelope.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "1c1aa5851c76cc7964f65bec6915d2da534e8bb83b0341d470d9c065ea093df3", "generation_seconds_batch": 14.66733354202006, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6669552326202393, "target_z": 2.1406249476030004, "ce": 6.308549880981445, "genuineness": 3.6851859092712402, "identity": 0.3678382635116577, "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_NEWS_reg/d2e1e854a388d1e1db1cf96c3328ba19d06bd3d89bdf749d1d1ae643c07c8268.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "d301637e3f342df6eafda02cc42a5fc1b43e4a4bce2250be3e9221accf42566b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9fc3213ea569c709d379137b5171373eef9568619a6cd2c22d879b9d8e147244", "generation_seconds_batch": 20.082217796938494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6359735727310181, "target_z": 0.26953124873233064, "ce": 6.213237762451172, "genuineness": 3.7918784618377686, "identity": 0.3428667187690735, "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_NEWS_reg/d301637e3f342df6eafda02cc42a5fc1b43e4a4bce2250be3e9221accf42566b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "d389babd7e9e3b73ec5cecc1233de4ba9c42a17ccd6fea57539a9f2f910adee1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "45e8dd48109d84cebfb1ba50ede546bdc5d10108b46e146788a648aaebe9a82b", "generation_seconds_batch": 20.19785993604455, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3311867713928223, "target_z": 1.5312501385985156, "ce": 6.38003396987915, "genuineness": 4.118232727050781, "identity": 0.34481340646743774, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/d389babd7e9e3b73ec5cecc1233de4ba9c42a17ccd6fea57539a9f2f910adee1.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "4261c5613bf737032bf78f4433ee0a57a0208a216789349ba66bb30be13de2af", "generation_seconds_batch": 13.90252202108968, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4474142789840698, "target_z": 1.7421875059157903, "ce": 5.895530700683594, "genuineness": 4.482243061065674, "identity": 0.5746662616729736, "hypothesis": "J'ai créé numero sur l'enveloppe. Le bien and verify the two dermo a moment.", "wer": 0.7272727272727273, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/d3fab0ead1fc9dcf899f59bddb0c9d4b54373ba1829b6dc49f77fc4027ce1740.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "d4188dd300201cdea71a94adac28e1198b38a9e7cd1e5839e3dc01f161c62d5c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "268c2922bb147960580aeb4772b26277968f5930ef92265db2350ff25d6887a3", "generation_seconds_batch": 14.02374214399606, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5595647096633911, "target_z": 0.13085939084586687, "ce": 6.41180419921875, "genuineness": 4.344170093536377, "identity": 0.3694632053375244, "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_NEWS_reg/d4188dd300201cdea71a94adac28e1198b38a9e7cd1e5839e3dc01f161c62d5c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "d48b0587ae2d35c93622bf89629ca121544b9a4cd0dc80cf3c26907630465c7c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "dc61f889133b41d10233412a2514a584b1c49fb5d01df18583444713e1db6035", "generation_seconds_batch": 18.766539247939363, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0083322525024414, "target_z": 0.9453124061924681, "ce": 6.467403411865234, "genuineness": 4.281409740447998, "identity": 0.4420406222343445, "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_NEWS_reg/d48b0587ae2d35c93622bf89629ca121544b9a4cd0dc80cf3c26907630465c7c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "d4b51aa1ad48104d33c20605be1826dc05e1c5ca5bdadbab39fc1f6ba13a92e7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0d011e24dff7ed6fbf381ba625f7718d911f01c90cf1bb1ce104ed1062fdfb10", "generation_seconds_batch": 21.66812719090376, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.140475034713745, "target_z": 3.0, "ce": 6.3641486167907715, "genuineness": 3.9927117824554443, "identity": 0.39584463834762573, "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_NEWS_reg/d4b51aa1ad48104d33c20605be1826dc05e1c5ca5bdadbab39fc1f6ba13a92e7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "d4e28e45dbb4123ae748129e34332d16561bf2a8b6c9a68d19852d6c93debedc", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "32e08fb88ce5b0df8a7facb89990eec254fcc4a4559a3b97d6e24f044e65e58a", "generation_seconds_batch": 20.19785993604455, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8232291340827942, "target_z": 0.6093750253533871, "ce": 6.38797664642334, "genuineness": 4.394378662109375, "identity": 0.3424680233001709, "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_NEWS_reg/d4e28e45dbb4123ae748129e34332d16561bf2a8b6c9a68d19852d6c93debedc.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "d5188f4bb95b0fda73e7fcb1ca8aeb4c09450beb72322d5120ad351f7d8a99a6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2e3ca0620abcc1bdc86e1f8e53024bb887ca827b7bf7500f72c98a6961d2ed40", "generation_seconds_batch": 18.782732837018557, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7960970401763916, "target_z": 2.375000040565419, "ce": 6.181467056274414, "genuineness": 4.061748504638672, "identity": 0.5254676342010498, "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_NEWS_reg/d5188f4bb95b0fda73e7fcb1ca8aeb4c09450beb72322d5120ad351f7d8a99a6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "d52268e2f60e5f75fe150607aec75e653401519477322282ebf08d5a183661f2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b98c966dc3fd410774820949e8c0b5f110f478913e76ee08d8a08458082ecbee", "generation_seconds_batch": 18.827360568917356, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9760468602180481, "target_z": 0.8867187411263145, "ce": 6.1894097328186035, "genuineness": 4.7081804275512695, "identity": 0.7353507280349731, "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_NEWS_reg/d52268e2f60e5f75fe150607aec75e653401519477322282ebf08d5a183661f2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "d52f057c1e5eb9b0796575bb247c5896f73c8cc7d9a250cc1a2ea2fb512add01", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "21c8991d7d75e9dd3d5e696c8acd4639c11474405bdc7abdcc47ca85a9942699", "generation_seconds_batch": 14.041933526983485, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.352710247039795, "target_z": 1.5703123656270488, "ce": 5.990842819213867, "genuineness": 4.030367851257324, "identity": 0.5338014364242554, "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_NEWS_reg/d52f057c1e5eb9b0796575bb247c5896f73c8cc7d9a250cc1a2ea2fb512add01.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "d5458cd52182ce76fa50c2b51e8cb80a6e83e6a57f595cb706406ffff2ce9c40", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "2c5d38dfbec7c22f55194263c0937337f0117447b9c40befc82d7b093adb0a21", "generation_seconds_batch": 20.533045855001546, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8210767507553101, "target_z": 0.605468737745863, "ce": 6.467403411865234, "genuineness": 4.306513786315918, "identity": 0.27635180950164795, "hypothesis": "I found the little box. It was beside the old window. We can look inside together. What", "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_NEWS_reg/d5458cd52182ce76fa50c2b51e8cb80a6e83e6a57f595cb706406ffff2ce9c40.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": "d569e96f5856fd138d0f6285aab2a15eb34e580c79cda33f226eb2414bf6023a", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Neem ze mee als je weggaat.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Natural clear speech.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Natural clear speech.) [3.2 seconds duration] Neem ze mee als je weggaat.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "22517db10934ec8a1ba84bdd113eb3e1e45f8f3fedf4d55a46cbd9fe541d4400", "generation_seconds_batch": 14.009174822014757, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0974278450012207, "target_z": 2.9218751132451284, "ce": 6.23706579208374, "genuineness": 4.218649387359619, "identity": 0.5137033462524414, "hypothesis": "De sleutels liggen op de keukentafel. Ik keek in de Larama 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_NEWS_reg/d569e96f5856fd138d0f6285aab2a15eb34e580c79cda33f226eb2414bf6023a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "d6087004896ffec6542e020ee588ff6703a156f3f52d876d60aace8ecb1bdd7f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "257f232d1e0f5a26afa8beed2d153904773282449701756f2ed7e3043798de58", "generation_seconds_batch": 18.830347500974312, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8921047449111938, "target_z": 0.7343750388751934, "ce": 6.117925643920898, "genuineness": 4.306513786315918, "identity": 0.6661278009414673, "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_NEWS_reg/d6087004896ffec6542e020ee588ff6703a156f3f52d876d60aace8ecb1bdd7f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "d6389d24c8c004e5e9898f9d76a267bc3a7301667c7d8e21143ccee58e5452cb", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4b34908d51d46afdc2ef69cd69d4938226f4e53296a4c601a463155a0564b96e", "generation_seconds_batch": 20.976721537066624, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.873582124710083, "target_z": 2.5156250963428706, "ce": 6.292664527893066, "genuineness": 4.243753433227539, "identity": 0.7276453375816345, "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_NEWS_reg/d6389d24c8c004e5e9898f9d76a267bc3a7301667c7d8e21143ccee58e5452cb.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "d6714d4f785e79f6bd7d373928c48a934b25e52764b8f42a5e5d1192b06adae8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "10da2271a7a19d31abc1ef23394a7192018427ec95fee6e66dc1c7bbae2d987d", "generation_seconds_batch": 15.872097639017738, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.140475034713745, "target_z": 3.0, "ce": 6.356205940246582, "genuineness": 3.8922953605651855, "identity": 0.512453019618988, "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_NEWS_reg/d6714d4f785e79f6bd7d373928c48a934b25e52764b8f42a5e5d1192b06adae8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "d707ac35c21eaeb64b0d61b31a699bcbe8b4489137036896b88353f75874c3bf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "b5ee462b97f91b7fe7d5cce523e8f722854070b580ac814bbfe024e29d76cae4", "generation_seconds_batch": 19.600245298002847, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7694200277328491, "target_z": 0.5117187005608953, "ce": 6.403861999511719, "genuineness": 4.18726921081543, "identity": 0.2922821640968323, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/d707ac35c21eaeb64b0d61b31a699bcbe8b4489137036896b88353f75874c3bf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "d70e99c759bea0b7f2dd0275d1685e1c1e7f832659dfe10edef9379ef3555f61", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2c3602924a736da69d46fb54178069f3190fe6a64781bd7ff1080012d1d2f125", "generation_seconds_batch": 20.423111359938048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152985095977783, "target_z": 2.0468749104180324, "ce": 6.340320587158203, "genuineness": 4.111956596374512, "identity": 0.45427438616752625, "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_NEWS_reg/d70e99c759bea0b7f2dd0275d1685e1c1e7f832659dfe10edef9379ef3555f61.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "d74a4e9812aaaf4deae62a8c787bb4d979cf8fe21b1d2a51f3297ac43f0eac93", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.0, "cfg": 1.25, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2eba1a36150ee1ff84eb83695f4bef82e1662fdd1463a92fb229b5756063df8e", "generation_seconds_batch": 15.069717325037345, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8791905641555786, "target_z": 0.7109375295789515, "ce": 6.300607204437256, "genuineness": 4.545003414154053, "identity": 0.6033211946487427, "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_NEWS_reg/d74a4e9812aaaf4deae62a8c787bb4d979cf8fe21b1d2a51f3297ac43f0eac93.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "d75b90b9a7f64b24ee650d94a57794b7482be81172b4c30f188673f0501c4e07", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "690a49fc21f4bc971e63078d846d879dfc08961ca4f1e6df921491023f4cef75", "generation_seconds_batch": 19.571819069911726, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4258906841278076, "target_z": 1.7031250625383545, "ce": 6.316492557525635, "genuineness": 4.432034492492676, "identity": 0.3597804009914398, "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_NEWS_reg/d75b90b9a7f64b24ee650d94a57794b7482be81172b4c30f188673f0501c4e07.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "d78a8b5a9ae94dded667f63e7f7a67fcaf1c2d94c010a76f5fa78c07fab42018", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4c1ee23cddb5117d6fa8bba5206bbc4884c03ec6a033d99a18bb810951507887", "generation_seconds_batch": 19.983958607073873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5821645259857178, "target_z": 0.1718750321142902, "ce": 6.2767791748046875, "genuineness": 4.419482707977295, "identity": 0.2787304222583771, "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_NEWS_reg/d78a8b5a9ae94dded667f63e7f7a67fcaf1c2d94c010a76f5fa78c07fab42018.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "d79d5bed73a23f0c60ac489d6ee420b56bc890a400a8e61533b3e91e2c2816a2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "72a8c5369d845f67dcdaa96c1c44e1df73786716553b6dc6220145873d6d44dc", "generation_seconds_batch": 20.131832786952145, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7780295014381409, "target_z": 0.5273437428165404, "ce": 6.268836498260498, "genuineness": 4.557555198669434, "identity": 0.5619996190071106, "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_NEWS_reg/d79d5bed73a23f0c60ac489d6ee420b56bc890a400a8e61533b3e91e2c2816a2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "d7dbb96b70d90df1ea73a1e7e35c2b0a182f10a26d3cab9e1c322226625cda61", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "58c21340c5fb16b53d2ebd290065dab7c3aeff023f84f9ee39e588ea33596d3a", "generation_seconds_batch": 18.833007132983766, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.174912929534912, "target_z": 3.06250016902258, "ce": 6.141753673553467, "genuineness": 4.193545341491699, "identity": 0.7538392543792725, "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_NEWS_reg/d7dbb96b70d90df1ea73a1e7e35c2b0a182f10a26d3cab9e1c322226625cda61.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": "d7f9b206f5c3a25c5ba0b7e0a0209411d0fb17a82b7e83bc63031dfb231660e8", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The keys are on the kitchen table.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Take them with you when you leave.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] The keys are on the kitchen table.\n(Natural clear speech.) [3.8 seconds duration] I checked the drawer... then looked beside the window.\n(Natural clear speech.) [3.2 seconds duration] Take them with you when you leave.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "4a254ebcd973cea0ef3b81085944cca2ebbea4a57c26b7228fae18dd724291b4", "generation_seconds_batch": 14.663563595036976, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1977401971817017, "target_z": 1.2890624704210485, "ce": 6.419747352600098, "genuineness": 4.256305694580078, "identity": 0.3142457902431488, "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_NEWS_reg/d7f9b206f5c3a25c5ba0b7e0a0209411d0fb17a82b7e83bc63031dfb231660e8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "d83735de33ff3d74364d035c74396dc66f35c992bb3361c813f3015078febf3c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "af82e17fa39b0f3e8d809e8f233f4fc1bbf4f33979faf02ab72fd8e3f87d79af", "generation_seconds_batch": 20.082217796938494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3441009521484375, "target_z": 1.5546876478947576, "ce": 6.197352409362793, "genuineness": 3.7416701316833496, "identity": 0.22792357206344604, "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_NEWS_reg/d83735de33ff3d74364d035c74396dc66f35c992bb3361c813f3015078febf3c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "d97ee32666d3e7629e989d0951ccd58aff046164277137895347cf8c31028d92", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "00f2a4fa5e9bfc2561530af05a0cfd710c4bba40fea1800d26ea2225d81d77e1", "generation_seconds_batch": 19.014288213918917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.3895280659198761, "target_z": -0.17773435535112508, "ce": 6.427689552307129, "genuineness": 4.482243061065674, "identity": 0.44735443592071533, "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_NEWS_reg/d97ee32666d3e7629e989d0951ccd58aff046164277137895347cf8c31028d92.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "d9921a8791331b71968f7142dcb344378e29b2c6692c86e437ba2a182036a2b4", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2cb7505bfbd89661e372345267eeb1b96a3fb01bd58f131b040d74701ecfeeab", "generation_seconds_batch": 14.665501038078219, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7866389751434326, "target_z": 0.5429687850721854, "ce": 6.252951145172119, "genuineness": 4.21237325668335, "identity": 0.30506202578544617, "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_NEWS_reg/d9921a8791331b71968f7142dcb344378e29b2c6692c86e437ba2a182036a2b4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "d9b50d16702cfa13c827c3f6a06262f8e243121ec463661591f1baef5cebdac0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "8e1095d8ebe5f1ff8665efe8d0d1cd83d118831947000e97bc2c9ba9c9e462b7", "generation_seconds_batch": 18.87938659498468, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4388048648834229, "target_z": 1.7265625718345965, "ce": 6.213237762451172, "genuineness": 4.344170093536377, "identity": 0.7206146717071533, "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_NEWS_reg/d9b50d16702cfa13c827c3f6a06262f8e243121ec463661591f1baef5cebdac0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "d9bde9db49d263f0dafd12b963fe535bc799d0cd7dd8601882f84d641ca8160f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6e49236d512fc679d340d8c5625b4f00521c5ecf60a1c6a07e4684affe814dc9", "generation_seconds_batch": 15.394675730960444, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6411268711090088, "target_z": 2.0937499290105164, "ce": 5.935244083404541, "genuineness": 3.8169827461242676, "identity": 0.5385147333145142, "hypothesis": "La porte est encore ouverte. J'ai posé ton manteau sur la chaise. Dis-moi quand tu es prêt. Jandy.", "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_NEWS_reg/d9bde9db49d263f0dafd12b963fe535bc799d0cd7dd8601882f84d641ca8160f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "da4045b4d016210bb098f6cc898c9654c61cd1f66952376de60676b0ca7a4975", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e645c43fbe6acbf8de67d33801ab6495a176ca438a55d6416876781cf16292da", "generation_seconds_batch": 20.372638594009914, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8133158683776855, "target_z": 2.4062499087278066, "ce": 6.221180438995361, "genuineness": 3.6663577556610107, "identity": 0.3576817512512207, "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_NEWS_reg/da4045b4d016210bb098f6cc898c9654c61cd1f66952376de60676b0ca7a4975.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "daa474cf2e3f11f5267fc97225a537230da971a9e1797f2c71ef7af22757bc3d", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6f1986240062a9a5cf1b8fe03a1b82d9b5349f4488ebb860889f08948747af66", "generation_seconds_batch": 20.18418435100466, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7242204546928406, "target_z": 0.4296875261984999, "ce": 6.24500846862793, "genuineness": 4.331617832183838, "identity": 0.5986390709877014, "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_NEWS_reg/daa474cf2e3f11f5267fc97225a537230da971a9e1797f2c71ef7af22757bc3d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "daf4762a8cb5f7b432adae42e574f90c212af75c82269578a0936829ec77629c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b31bc82a2420db1d087429f66e33a2a8e2a1282c0787ce2f31ba74f618e3201f", "generation_seconds_batch": 15.409564901026897, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0199427604675293, "target_z": 2.7812500574676773, "ce": 6.348263263702393, "genuineness": 4.218649387359619, "identity": 0.5145283341407776, "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_NEWS_reg/daf4762a8cb5f7b432adae42e574f90c212af75c82269578a0936829ec77629c.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": "dafd68a8865426df89d4e332efa5eeb2228109d3e55972cdaa8c75819374ca16", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1701, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2c2e13bafec303a60ad5a960792f753c5627a72a22ddd3cb38b1d5c9f86d66ef", "generation_seconds_batch": 14.552804628037848, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7788782119750977, "target_z": 2.3437501724030314, "ce": 6.229123115539551, "genuineness": 4.143336772918701, "identity": 0.7366518378257751, "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_NEWS_reg/dafd68a8865426df89d4e332efa5eeb2228109d3e55972cdaa8c75819374ca16.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "db182a13c55bd14ccf4523e4a7329f7136a3de0ec8b1fe974bc3d20ee267383c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b895986f542c6dea05894c6a20652ba15d4344b0b90796314241c4d49710106d", "generation_seconds_batch": 13.991706050001085, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1202552318572998, "target_z": 1.1484376309924995, "ce": 6.284721851348877, "genuineness": 3.9048471450805664, "identity": 0.022319171577692032, "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_NEWS_reg/db182a13c55bd14ccf4523e4a7329f7136a3de0ec8b1fe974bc3d20ee267383c.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 127, "seconds_generated": 10.16, "mp3_sha256": "9d8cda047eaffb726f96a68037d89a41fa44cfaf9788fdfa92f4f2551f9860ee", "generation_seconds_batch": 13.868217114009894, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8921047449111938, "target_z": 0.7343750388751934, "ce": 6.308549880981445, "genuineness": 3.9676074981689453, "identity": 0.30865156650543213, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/db192910e9ae67de251fb2a990940eab881f66f627c78b50b638ce61f5a63679.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "dbd2a43d688f1e1a164774d51ffdf9d6dcbc13a37f57ee3e6a0e3c50615d3d46", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "96ac85a7d9a900efd2b79c62ceead65c964b51127bbb2ade2ba857df84a7493e", "generation_seconds_batch": 20.816095154033974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4861568212509155, "target_z": 1.812500033804516, "ce": 6.078212261199951, "genuineness": 4.406930446624756, "identity": 0.6821538209915161, "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_NEWS_reg/dbd2a43d688f1e1a164774d51ffdf9d6dcbc13a37f57ee3e6a0e3c50615d3d46.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "dc6b8c47caad4b0973a94ea7bd27d64d6db94612cc4cf264aea61d55cf301625", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9e1410997c749c29279320d83da0273473c7853e437264e3a36847233bc28b72", "generation_seconds_batch": 18.883292009006254, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.780181884765625, "target_z": 0.5312500304240644, "ce": 6.24500846862793, "genuineness": 4.494794845581055, "identity": 0.5558476448059082, "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_NEWS_reg/dc6b8c47caad4b0973a94ea7bd27d64d6db94612cc4cf264aea61d55cf301625.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "dc7500bf971d3e5157d3c864ab57af66bcef5529c8a276ba5663db25f6327596", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b78dac2ef1937b32b20c5bf3edba77aa7c01a9cb9d2289e3201cd0ab09f86e8f", "generation_seconds_batch": 20.98670075798873, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9480661153793335, "target_z": 0.8359374349263067, "ce": 6.24500846862793, "genuineness": 4.532451152801514, "identity": 0.5820199251174927, "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_NEWS_reg/dc7500bf971d3e5157d3c864ab57af66bcef5529c8a276ba5663db25f6327596.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "dcdd5978ef5be5c81b40bd531b5811816e4eb800c53565f3421f23d10d299eab", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "692d0ee1e2f9ef72af2f764c54329e8aaf8efd0120e475f858ab89856c9a0214", "generation_seconds_batch": 21.049558890983462, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.38003396987915, "genuineness": 4.080576419830322, "identity": 0.49964216351509094, "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_NEWS_reg/dcdd5978ef5be5c81b40bd531b5811816e4eb800c53565f3421f23d10d299eab.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": "dd235bf07b4cec5919130aa7ed29160baf69aea9388cacca8e6f344382223760", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Je bericht kwam nadat ik vertrokken was.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik las het twee keer... en legde mijn telefoon weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "86549a0ade957a394b6a962e2a581a04f401c2586bdc740646737eab39791081", "generation_seconds_batch": 20.62028798495885, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9855048656463623, "target_z": 2.7187498884450974, "ce": 6.006728172302246, "genuineness": 4.030367851257324, "identity": 0.42882221937179565, "hypothesis": "Je bericht kwam nadat ik vertrokken was. Ik las het twee keer en legde mijn telefoon weg. Zeg maar waar Deelweg kunnen praten.", "wer": 0.08695652173913043, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/dd235bf07b4cec5919130aa7ed29160baf69aea9388cacca8e6f344382223760.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": "dd3b73676081bffa9535e1236098e4228cff9500ce95bf7429321eb2b6718d80", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Deine Nachricht kam, als ich schon weg war.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich las sie zweimal... dann legte ich das Handy weg.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2a125e286785ea5e8e0f95028b1fc797f210e9a8cc0d2e7203614f88f603cac6", "generation_seconds_batch": 20.000740652903914, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9911134243011475, "target_z": 0.9140625380300805, "ce": 6.356205940246582, "genuineness": 4.306513786315918, "identity": 0.5271458625793457, "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_NEWS_reg/dd3b73676081bffa9535e1236098e4228cff9500ce95bf7429321eb2b6718d80.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "dd41c6e71b351042a7100c65ecbd9ea5f091b8daaa920a3f032b59c0ea719914", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9352c42cb5ced9f0dcfebf3b324a7fa9e298da44ec7bbbbb9b4797848fc7f68b", "generation_seconds_batch": 18.757836426026188, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9855048656463623, "target_z": 2.7187498884450974, "ce": 6.229123115539551, "genuineness": 4.218649387359619, "identity": 0.8558505177497864, "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_NEWS_reg/dd41c6e71b351042a7100c65ecbd9ea5f091b8daaa920a3f032b59c0ea719914.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "dd4634cee5564e6d2962da919126db689f3ac5827cf8d637b6e4b34035f0b2ee", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e28fdea9785236da2aa789d40a72ad50d04139433ad8bba45af60dc4f313f365", "generation_seconds_batch": 15.576549438992515, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.062326908111572, "genuineness": 4.168440818786621, "identity": 0.5148499608039856, "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_NEWS_reg/dd4634cee5564e6d2962da919126db689f3ac5827cf8d637b6e4b34035f0b2ee.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "dd6c58db878dcdfc0e6bfe6b78aba81a83c1ac48fdbaf4997f7b97597efec5c9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "ea152546e5c9f3bc21f8d956586dc3248dcee0f873c10ef54d740a2b5117d69c", "generation_seconds_batch": 18.811133375042118, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8834952712059021, "target_z": 0.7187499966195484, "ce": 6.133810997009277, "genuineness": 4.494794845581055, "identity": 0.7025389671325684, "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_NEWS_reg/dd6c58db878dcdfc0e6bfe6b78aba81a83c1ac48fdbaf4997f7b97597efec5c9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "dd88c0dc55b50e4c7c5ebd362d4cdb28d028a76dfbf86e691125f2ce8efe9d81", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "71440c4b461e0ff57919055f2be113f8a0842b77cd45b08a4dbb521538be25a0", "generation_seconds_batch": 19.61942932603415, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9437614679336548, "target_z": 0.8281250760601611, "ce": 6.284721851348877, "genuineness": 4.645420074462891, "identity": 0.6937451958656311, "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_NEWS_reg/dd88c0dc55b50e4c7c5ebd362d4cdb28d028a76dfbf86e691125f2ce8efe9d81.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": "ddca9a00d460615c2e3a65b2e1fa1a0d88d6b3d5bbac128fdb00dfb7f344c03b", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Jemand hat den kleinen blauen Stuhl umgestellt.\n(Natural clear speech.) [3.8 seconds duration] Er stand hier... neben dem hohen Bücherregal.\n(Natural clear speech.) [3.2 seconds duration] Wir stellen ihn nach dem Essen zurück.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b6b24ce0e155f34a68727cc12fbd41757659aa606b53b9613c8c975d9ae334c8", "generation_seconds_batch": 13.942868506070226, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9510672092437744, "target_z": 2.656250152120322, "ce": 6.403861999511719, "genuineness": 4.2751336097717285, "identity": 0.5008255839347839, "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_NEWS_reg/ddca9a00d460615c2e3a65b2e1fa1a0d88d6b3d5bbac128fdb00dfb7f344c03b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "df299d63bf7d3fea61c90a2e9bddda6a30150e3cb12fed05a135ce435a357d53", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "41e49a2d411a4e8e59b87e33226e0994728d648f17967ed416acd6f92ffe8522", "generation_seconds_batch": 20.597606427036226, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.157693862915039, "target_z": 3.0312498681623876, "ce": 6.1894097328186035, "genuineness": 3.948779582977295, "identity": 0.7861442565917969, "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_NEWS_reg/df299d63bf7d3fea61c90a2e9bddda6a30150e3cb12fed05a135ce435a357d53.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": "df8cddecb7f4eb5aab9cbf176e2c2f9909c2cc2a01eb1644c25c184f55b087e4", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "6fce530d4d3a46ad755d75e28ac93c8d1e539db31d4552875f8cc44274c279d7", "generation_seconds_batch": 19.61420097493101, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7702686786651611, "target_z": 2.3281250219729355, "ce": 6.125868320465088, "genuineness": 4.042920112609863, "identity": 0.7556660175323486, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/df8cddecb7f4eb5aab9cbf176e2c2f9909c2cc2a01eb1644c25c184f55b087e4.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "dfa331d02d5d3b5e943e1404d8b84c6af6652ea84ba35af7488c4755750a8d4e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a91afefc68dbd31ee01981c12ccfd270a7058ab2dc5021ed9fdcc51afae3b1d8", "generation_seconds_batch": 16.39072722697165, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9868086576461792, "target_z": 0.9062499628150325, "ce": 6.109982967376709, "genuineness": 4.632867813110352, "identity": 0.6279290914535522, "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_NEWS_reg/dfa331d02d5d3b5e943e1404d8b84c6af6652ea84ba35af7488c4755750a8d4e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "e0027a860eca1a50ca882947e5a8026e4d70bf3937e80356e23e89f7b22f5b1c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "7c6e1fd7e2863250c10401a83e096faa002431992a8723e0ee3de8b3cac112fe", "generation_seconds_batch": 18.91904967406299, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2192637920379639, "target_z": 1.3281249137984843, "ce": 6.070269584655762, "genuineness": 4.118232727050781, "identity": 0.6382293701171875, "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_NEWS_reg/e0027a860eca1a50ca882947e5a8026e4d70bf3937e80356e23e89f7b22f5b1c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "e090f4111213b7cef0c8326ba0b701d28a12bf329e97dbfa3ae4ee1856687df3", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c3beddda80eeebc348eaa8d6372c5b6279559600e15b83262fdc3824ba19cb73", "generation_seconds_batch": 19.215223321923986, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.55503249168396, "target_z": 1.9375001555007736, "ce": 6.046441555023193, "genuineness": 4.093128204345703, "identity": 0.3723986744880676, "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_NEWS_reg/e090f4111213b7cef0c8326ba0b701d28a12bf329e97dbfa3ae4ee1856687df3.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "e19e8c64af93a126f2eb13b21d8876a29b9574016b22bdef71df1f9e686cc19a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "75fbca357231bf4b565bb75dedf2db76f63af38866df5e13aea36aca4b889615", "generation_seconds_batch": 20.15223007905297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2881394624710083, "target_z": 1.4531250354947418, "ce": 6.1894097328186035, "genuineness": 4.281409740447998, "identity": 0.6585947275161743, "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_NEWS_reg/e19e8c64af93a126f2eb13b21d8876a29b9574016b22bdef71df1f9e686cc19a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "e3371f81a902ee4556abe0d44eb6bd0a44ce13993948b47adb4e413d06bf7051", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "bdcd62ed3711ae3d7d88794034a9877fc9d3c6a7745e159f40af32ebf73bf357", "generation_seconds_batch": 21.045267187058926, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8649725914001465, "target_z": 2.499999945912774, "ce": 6.078212261199951, "genuineness": 3.9801597595214844, "identity": 0.7087584733963013, "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_NEWS_reg/e3371f81a902ee4556abe0d44eb6bd0a44ce13993948b47adb4e413d06bf7051.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "e3568ca48e270845844d60784cd20a746d7663bf596310813e2f45c1cda6fc1f", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 128, "seconds_generated": 10.24, "mp3_sha256": "fad56da568d5a8464c8ab1f8bee9b8f72b8f43d7baa65e407ae7dd8cb0c1912d", "generation_seconds_batch": 14.591332069016062, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.115950345993042, "target_z": 1.140624839428549, "ce": 6.078212261199951, "genuineness": 4.494794845581055, "identity": 0.7328097224235535, "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_NEWS_reg/e3568ca48e270845844d60784cd20a746d7663bf596310813e2f45c1cda6fc1f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "e36ada6d7725404dc50d57a2e67a6e1687acb6f5bbfd47535683aea632568443", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "a7f807b50d628d5b7d4fa4b03b0c12c0d361ceb12bfb2c9f93d58909bf4cceb7", "generation_seconds_batch": 19.02778282703366, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.268836498260498, "genuineness": 4.457139015197754, "identity": 0.6979376673698425, "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_NEWS_reg/e36ada6d7725404dc50d57a2e67a6e1687acb6f5bbfd47535683aea632568443.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": "e3f35ee72a1ec04a19ccd2b443da0d814556569a61b2a19baae2ad61fe917155", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Natural clear speech.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Natural clear speech.) [3.2 seconds duration] Nimm sie mit, wenn du gehst.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "a103e6a85cb1ffbe899851d41262b77a9cb2a734f60e13539cd7319fffd84942", "generation_seconds_batch": 15.491098306025378, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1589977741241455, "target_z": 1.2187501588812253, "ce": 6.308549880981445, "genuineness": 4.381826400756836, "identity": 0.493010014295578, "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_NEWS_reg/e3f35ee72a1ec04a19ccd2b443da0d814556569a61b2a19baae2ad61fe917155.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "e43c1fe2b85e979d492e86c82a1c07cc7312cfe996c2aa05bfc1716a1a3c020b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "1e3e018bdc02a6a69d31920ca861de4bacfa502fe7e0b91c468efd2477192ec7", "generation_seconds_batch": 18.757836426026188, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.619830846786499, "target_z": 0.24023436211202828, "ce": 6.268836498260498, "genuineness": 4.457139015197754, "identity": 0.7720541954040527, "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_NEWS_reg/e43c1fe2b85e979d492e86c82a1c07cc7312cfe996c2aa05bfc1716a1a3c020b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "e4e492f03660d75cf97057592d8e7c1f7e1eab990b28c23b03ce9a0896d5ff2e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9dd77244606faeb93c8ee22b415623525404dd2249e639e522644f8e56325a48", "generation_seconds_batch": 18.850683488999493, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3570151329040527, "target_z": 1.5781251571909993, "ce": 6.181467056274414, "genuineness": 4.331617832183838, "identity": 0.67784184217453, "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_NEWS_reg/e4e492f03660d75cf97057592d8e7c1f7e1eab990b28c23b03ce9a0896d5ff2e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "e503da7cc4a429018ce28c48798563689faf23d0321f95a78c0d1b6aabbc9c83", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9c32f853242ec4babd1f8e1603f75052e94c48dba92572345edc184077304d2c", "generation_seconds_batch": 19.238005773047917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3742339611053467, "target_z": 1.609375025353387, "ce": 6.133810997009277, "genuineness": 4.256305694580078, "identity": 0.6532389521598816, "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_NEWS_reg/e503da7cc4a429018ce28c48798563689faf23d0321f95a78c0d1b6aabbc9c83.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "e619bc2a5141ee3b0bd5276638e70c778c1e70e6308cf5e893da70b9d56135bb", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "eb953333a182d828ba4fe2367499183e579b9d46fc8f61364252ffe743726372", "generation_seconds_batch": 20.95940620091278, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7530498504638672, "target_z": 2.296875153810548, "ce": 6.427689552307129, "genuineness": 4.093128204345703, "identity": 0.473001629114151, "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_NEWS_reg/e619bc2a5141ee3b0bd5276638e70c778c1e70e6308cf5e893da70b9d56135bb.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "e61e56ae8107afbcb20fd98d47a4c2307782f347bbda96cbf5eae612a543ce45", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2f04281a4f529fe109a49d2349fecb28d359c1a6af8de088e1a836699a2ab3d2", "generation_seconds_batch": 18.842281606979668, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5649455785751343, "target_z": 0.1406249476030002, "ce": 6.499174118041992, "genuineness": 4.607763767242432, "identity": 0.4519197940826416, "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_NEWS_reg/e61e56ae8107afbcb20fd98d47a4c2307782f347bbda96cbf5eae612a543ce45.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "e637d89a66f68833c2ae8e4dde4448c2f69f8a5e359af2dd9155bc007e57eb1a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b5ab724eeb5d9f4d828287b88b9df0b9fceabc906643f075807ae636aae24a80", "generation_seconds_batch": 20.670553746982478, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.025551199913025, "target_z": 0.9765624907037581, "ce": 6.23706579208374, "genuineness": 4.432034492492676, "identity": 0.5581870675086975, "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_NEWS_reg/e637d89a66f68833c2ae8e4dde4448c2f69f8a5e359af2dd9155bc007e57eb1a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "e6eb483df2e3b5d532cdac1cbefa8bfb27598c2d055f73400d4332f21e9f6b3a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "018ba196fb839f2e8e6c734a94483550c477d5512be3d6a5906df4267ed4fbe3", "generation_seconds_batch": 19.683334450935945, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.046441555023193, "genuineness": 3.9801597595214844, "identity": 0.5495141744613647, "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_NEWS_reg/e6eb483df2e3b5d532cdac1cbefa8bfb27598c2d055f73400d4332f21e9f6b3a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "e6ec922fe4b6a900245280c1a308f2adf97cf6362e0486b6c50459a2a0a9e7bf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "3ec083cf5e7e8baacb635e2687bed026d7b16d069ad9a37341271f93e0427ba9", "generation_seconds_batch": 22.236294489935972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9338481426239014, "target_z": 2.62499985126013, "ce": 6.086154937744141, "genuineness": 3.6412534713745117, "identity": 0.5216035842895508, "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_NEWS_reg/e6ec922fe4b6a900245280c1a308f2adf97cf6362e0486b6c50459a2a0a9e7bf.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "5fb3ddffed3f88ca7d90ffa44189593ef1af75c761dbb0aac446cc030138f10a", "generation_seconds_batch": 14.339589474955574, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3742339611053467, "target_z": 1.609375025353387, "ce": 6.427689552307129, "genuineness": 4.457139015197754, "identity": 0.31555575132369995, "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_NEWS_reg/e6f1e80cc6606ef00fdc322d5829b94fd65ad80a74ab015f88321e5dc194804c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "e824f4e6fa119aefb89aa4e889a2504e9093e853b8bb538df9212c982029b144", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "bf202a9a7ffd499a37beebe1ac56e50bb1e1ee03de04dc724a026b69a5831f5e", "generation_seconds_batch": 15.591454515000805, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8275338411331177, "target_z": 0.6171874923939838, "ce": 6.2767791748046875, "genuineness": 4.118232727050781, "identity": 0.2543148398399353, "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_NEWS_reg/e824f4e6fa119aefb89aa4e889a2504e9093e853b8bb538df9212c982029b144.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "e87918f1b6a5f8bfedee5f33e1dd5913e6f452fd8b9fd235459eeb6d14291554", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "dd3cf270677bf484c1a958a573ffb0543097a02baf0e03243b80295148aa5044", "generation_seconds_batch": 20.546833553002216, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8047065734863281, "target_z": 2.3906251909955154, "ce": 6.252951145172119, "genuineness": 4.2249250411987305, "identity": 0.4308921694755554, "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_NEWS_reg/e87918f1b6a5f8bfedee5f33e1dd5913e6f452fd8b9fd235459eeb6d14291554.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "e8a68fdb0ee7ade7c243b2b346965490414231e20db4626a16841d3d42e81e53", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4c0fb094112e963d195d804530da51a6395e889533d60565b39bc00f6f2b7689", "generation_seconds_batch": 15.409564901026897, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.3384923934936523, "target_z": 3.3593749983097743, "ce": 6.4515180587768555, "genuineness": 3.6851859092712402, "identity": 0.45752793550491333, "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_NEWS_reg/e8a68fdb0ee7ade7c243b2b346965490414231e20db4626a16841d3d42e81e53.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "e91196b4d8ef7545abd4b49a87bd27729b3432de813fb2a4c7a74dbfc4ac55c5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4154a3d8b79814d5878dd03c9b38b5fe38884b37f1e39067a3c0ebeef9cb4cd1", "generation_seconds_batch": 20.131832786952145, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.537813425064087, "target_z": 1.9062498546405813, "ce": 6.165581703186035, "genuineness": 4.168440818786621, "identity": 0.4811338782310486, "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_NEWS_reg/e91196b4d8ef7545abd4b49a87bd27729b3432de813fb2a4c7a74dbfc4ac55c5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "e957d8caf18a496fce960a1832b98aa8264b733a57e3d1a2676a6754177f8205", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "aee5ea6f9039749823cfd54958c0dbd3565c206b548221969bd464c3df717f26", "generation_seconds_batch": 16.103394269011915, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.965285062789917, "target_z": 0.8671875194375968, "ce": 6.252951145172119, "genuineness": 4.482243061065674, "identity": 0.6311590671539307, "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_NEWS_reg/e957d8caf18a496fce960a1832b98aa8264b733a57e3d1a2676a6754177f8205.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "e9be2c2dff4f5fa0ff2c19d31e955603a123b332087c54118ffc4bfec9a22f6b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.5, "cfg": 3.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "6372a4579df3b2c64f4701d08287b7594195d3728c415868f67118d9c6b959cb", "generation_seconds_batch": 18.959711924078874, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2063496112823486, "target_z": 1.3046874045022423, "ce": 6.133810997009277, "genuineness": 4.432034492492676, "identity": 0.6784285306930542, "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_NEWS_reg/e9be2c2dff4f5fa0ff2c19d31e955603a123b332087c54118ffc4bfec9a22f6b.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "ed0ed1a8173b646a6a4d8ed96c7f2e0444ce6ee397d394f88bff8de08b5d4f9a", "generation_seconds_batch": 14.915260025998577, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6411268711090088, "target_z": 2.0937499290105164, "ce": 6.340320587158203, "genuineness": 4.494794845581055, "identity": 0.46637922525405884, "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_NEWS_reg/e9d266c3907611a53d0c5f7688a7e231bba29ead6f87b8ae8d29cb37ab2fae41.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "e9eb713d96c86630823c05808f90ea82586827d2c8dda0b074aaded87effb5be", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9a915bac4fc387769fbfb3bc90187812d5722e9c4ceab35420743f1617fdf0db", "generation_seconds_batch": 20.139590786071494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.839144229888916, "target_z": 2.4531249273202906, "ce": 6.0305562019348145, "genuineness": 4.256305694580078, "identity": 0.3874114751815796, "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_NEWS_reg/e9eb713d96c86630823c05808f90ea82586827d2c8dda0b074aaded87effb5be.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "e9ed4e93d28c961038865e4ad167e8d220d69f845535705a81c4b2072fdcd176", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1702, "scale": 1.0, "cfg": 2.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "3105ee38f4509b129c7aa0976fb2e1ca9c61d614e79b83de019103cb940fef34", "generation_seconds_batch": 19.50242937298026, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1762166023254395, "target_z": 1.250000027043613, "ce": 6.1020402908325195, "genuineness": 4.394378662109375, "identity": 0.7133520841598511, "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_NEWS_reg/e9ed4e93d28c961038865e4ad167e8d220d69f845535705a81c4b2072fdcd176.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "eaf4a65afd556b3e626fe9951ca193e391cc2d1019ea0b233ae48fdfd1c918ce", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d44f67ad45222228302f0c6892e8b6c0bcab6d7a3b5787f58e1b0119f8d01230", "generation_seconds_batch": 19.61908814799972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.197352409362793, "genuineness": 4.570107460021973, "identity": 0.7247418165206909, "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_NEWS_reg/eaf4a65afd556b3e626fe9951ca193e391cc2d1019ea0b233ae48fdfd1c918ce.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "eb0ecc597d6eb95d2f75180cc142a240a10863f063b79127455e565178ca35e7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "eb179e60d5c565ab4104beb6e3b7b97bd06491cefafd0a93a4286abe61478117", "generation_seconds_batch": 19.377071659080684, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2450921535491943, "target_z": 1.374999932390968, "ce": 6.300607204437256, "genuineness": 4.143336772918701, "identity": 0.4006904661655426, "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_NEWS_reg/eb0ecc597d6eb95d2f75180cc142a240a10863f063b79127455e565178ca35e7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "eb1f6600b7ae673f3506fa5c6bd79e03f919074cdb2db73882554ce961b260b0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d8397e7af504c986c38c956daee6a5da3c8ac858b1d5e263d39317cc18cfe36c", "generation_seconds_batch": 19.5266365269199, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.41643258929252625, "target_z": -0.12890624704210485, "ce": 6.324435234069824, "genuineness": 3.9425034523010254, "identity": 0.10805090516805649, "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_NEWS_reg/eb1f6600b7ae673f3506fa5c6bd79e03f919074cdb2db73882554ce961b260b0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "eb7760d25c207bacb8d899c61451a30311c1c948606bebe107bb00f3cb8ae725", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "5dc7809720d68dafa576bf41103738ac508cdcab9bb215fe219365007569b672", "generation_seconds_batch": 18.88016403396614, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.150388240814209, "target_z": 1.203125008451129, "ce": 6.181467056274414, "genuineness": 4.337893962860107, "identity": 0.36548906564712524, "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_NEWS_reg/eb7760d25c207bacb8d899c61451a30311c1c948606bebe107bb00f3cb8ae725.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": "eb89a1c0c086b8b8c2be67999af7406dbb030a5eb600e9207c7ff64f5c097669", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Schlüssel liegen auf dem Küchentisch.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich suchte in der Schublade... dann neben dem Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "57283ca60adba7111f3ee67e409f5e746d35a1b16ab5c50c77a4da8a641d9825", "generation_seconds_batch": 20.21829507104121, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.0113332271575928, "target_z": 2.765624907037581, "ce": 6.268836498260498, "genuineness": 3.898571252822876, "identity": 0.35499319434165955, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/eb89a1c0c086b8b8c2be67999af7406dbb030a5eb600e9207c7ff64f5c097669.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "eb973786f10b78d4dc03c832b44d54cb73ae6834dce0778aa2f46e34bd2b7d07", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "f10d75b8e6e209f298300453e7e259e6491d314592ea19ed4735caf1bf40f9ad", "generation_seconds_batch": 20.163828142918646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0513795614242554, "target_z": 1.023437509296242, "ce": 6.427689552307129, "genuineness": 4.256305694580078, "identity": 0.2867969870567322, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/eb973786f10b78d4dc03c832b44d54cb73ae6834dce0778aa2f46e34bd2b7d07.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "eba78b8095a56f387d626d39720d23873a04bf1e79295a92ec100c9fa47a97cb", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "e2b96175d2cc554f61df340f5f7415c7edae0a5743cf5a23a670752fbb39e4b5", "generation_seconds_batch": 20.522779140039347, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.460328459739685, "target_z": 1.7656250152120323, "ce": 6.149696350097656, "genuineness": 4.457139015197754, "identity": 0.6928333044052124, "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_NEWS_reg/eba78b8095a56f387d626d39720d23873a04bf1e79295a92ec100c9fa47a97cb.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "ebc56c2e067129d885fff11cd4dd4a017aaeddea006243e63d692f00b090f655", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "4f3ba73a12bf6a53a2ea6b1b6ed3472b647ffd18795428fe4dfdfa7619b546a1", "generation_seconds_batch": 18.87042517401278, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5636417865753174, "target_z": 1.953124873233065, "ce": 6.038498878479004, "genuineness": 4.061748504638672, "identity": 0.7458876371383667, "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_NEWS_reg/ebc56c2e067129d885fff11cd4dd4a017aaeddea006243e63d692f00b090f655.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "09e5d0a80e73e297148458f4b2d4adea5a4d5f5b902527ada36dae9c4c90eb61", "generation_seconds_batch": 13.889720431994647, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1374740600585938, "target_z": 1.1796874991548871, "ce": 6.1020402908325195, "genuineness": 4.325342178344727, "identity": 0.7774322032928467, "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_NEWS_reg/ebe89981a89e9506d02bacdfa583f9bfe1f04faa699193749ab6419de85c588f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "ec2ca79b78ebc3d14dc39200681843fe2ff43d044954d4eeb5482c630cf07ed2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c7c398a36764a6937d30c2ab88a3e028bd33e638e2c0caed76a9fdeecaff873c", "generation_seconds_batch": 19.54425967205316, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3053584098815918, "target_z": 1.4843751200060318, "ce": 6.141753673553467, "genuineness": 4.118232727050781, "identity": 0.6190598011016846, "hypothesis": "De deur staat nog open. Ik heb je jas op de stoel gelegd. Ik heb je jas op de stoel gelegd.", "wer": 0.42105263157894735, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/ec2ca79b78ebc3d14dc39200681843fe2ff43d044954d4eeb5482c630cf07ed2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "ec908821638f6bd0630a1cff28bc05151812856332b618767873d2f38f9f7aab", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d69a906f221fb77652133a4e81a16a8e1f7d62b920bacdfb78eb2f6618ebc183", "generation_seconds_batch": 14.27198091999162, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2537016868591309, "target_z": 1.3906250828210642, "ce": 6.197352409362793, "genuineness": 4.607763767242432, "identity": 0.5138168334960938, "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_NEWS_reg/ec908821638f6bd0630a1cff28bc05151812856332b618767873d2f38f9f7aab.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "ecc0e25471b8cda613c6dc718232edaf8a40802c3958c4b57381144438390024", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "687337c8d2d996ddb6aaf1da50dc2448f05e98649c3243a2fc329b1789568640", "generation_seconds_batch": 19.604444803087972, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3742339611053467, "target_z": 1.609375025353387, "ce": 6.221180438995361, "genuineness": 4.20609712600708, "identity": 0.559525728225708, "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_NEWS_reg/ecc0e25471b8cda613c6dc718232edaf8a40802c3958c4b57381144438390024.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "ecd76e74ec42afd7e4a3b15247104040f147af862040ec9db766776ef52972ae", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6061cc0d78547ff24e2a5d802bf4ed0e8a9f86d4ac8e0fad4eb2f26e5f667900", "generation_seconds_batch": 19.601447627996095, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6693351864814758, "target_z": 0.3300781116894718, "ce": 6.165581703186035, "genuineness": 4.582659721374512, "identity": 0.4764401316642761, "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_NEWS_reg/ecd76e74ec42afd7e4a3b15247104040f147af862040ec9db766776ef52972ae.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "ed4eeae645364957e5738edad67efe4b1235685617f47d68eaf4851387ba0118", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "f1a2ce51086e704d76d872f3b658bebe27671fd5f726cdd204cebd4cc77ab155", "generation_seconds_batch": 13.991706050001085, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7392870187759399, "target_z": 0.45703132310226585, "ce": 6.141753673553467, "genuineness": 4.005263805389404, "identity": 0.23444697260856628, "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_NEWS_reg/ed4eeae645364957e5738edad67efe4b1235685617f47d68eaf4851387ba0118.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "ed8a042b0eb765c9ac8fbe52853bc527416cd0c24ed714136e4389cc44274f0a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La porte est encore ouverte.\n(Natural clear speech.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand tu es prêt.", "seed": 1701, "scale": 0.75, "cfg": 2.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "d27d09722799dc485ca227224450379dc6ec6a4b56f1ff6109b6379d4aa54246", "generation_seconds_batch": 19.238005773047917, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4560236930847168, "target_z": 1.757812439996984, "ce": 6.125868320465088, "genuineness": 4.394378662109375, "identity": 0.7812305092811584, "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_NEWS_reg/ed8a042b0eb765c9ac8fbe52853bc527416cd0c24ed714136e4389cc44274f0a.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": "edd059a1ef9a4be107ffbf3f5ab2ef34490ae55009b77313307fcb78975cca10", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Escribí el número en el sobre.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Llámame cuando tengas un momento.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Escribí el número en el sobre.\n(Natural clear speech.) [3.8 seconds duration] Léelo bien... y comprueba las dos últimas cifras.\n(Natural clear speech.) [3.2 seconds duration] Llámame cuando tengas un momento.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "0be35eba54d1cd5dd93e320b505564ef80c2c03441f7a71174d133500175b7a5", "generation_seconds_batch": 15.709006855031475, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9338481426239014, "target_z": 2.62499985126013, "ce": 6.221180438995361, "genuineness": 3.9676074981689453, "identity": 0.7115738391876221, "hypothesis": "Escribí el número en el sobre, léele bien, y comprueba las dos últimas cifras. Llámame cuando tengas un momento.", "wer": 0.05263157894736842, "duration_error_s": 0.040000000000000924, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/edd059a1ef9a4be107ffbf3f5ab2ef34490ae55009b77313307fcb78975cca10.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": "ee5061cfbe0b8d4eb721dd8dd6536a934bd266d5b2d6b8e46709b9e8b7a7c032", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De sleutels liggen op de keukentafel.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik keek in de la... en daarna naast het raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 1.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5704242dd36fc121cf7f3832fd8b3eb9beeab51e99ff369c90417f79a6dada86", "generation_seconds_batch": 19.44807394396048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.165581703186035, "genuineness": 4.406930446624756, "identity": 0.7208252549171448, "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_NEWS_reg/ee5061cfbe0b8d4eb721dd8dd6536a934bd266d5b2d6b8e46709b9e8b7a7c032.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "eead913767c1f99b666e87dc10bd3be71e751d4397ff3a56d8265448e81a214c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "dd5e101d8b54092ee7d45fa12081fe95eaaf7ae170b238b533a365e33d33dfbc", "generation_seconds_batch": 19.524337145965546, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0018751621246338, "target_z": 0.9335936515443471, "ce": 6.109982967376709, "genuineness": 4.118232727050781, "identity": 0.4177430272102356, "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_NEWS_reg/eead913767c1f99b666e87dc10bd3be71e751d4397ff3a56d8265448e81a214c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "ef10eaafc6f26727daaf8c56b3c7536fcbfc5f1dd13501a9f6ddfa54f7c6b7fe", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1701, "scale": 0.5, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "cf1f0429c847545d9d38b1150c071d624a3e4334166c3d57bdbe5b742a2e44d9", "generation_seconds_batch": 19.58714419102762, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.133169412612915, "target_z": 1.1718751402887415, "ce": 6.173524379730225, "genuineness": 4.18099308013916, "identity": 0.5877363681793213, "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_NEWS_reg/ef10eaafc6f26727daaf8c56b3c7536fcbfc5f1dd13501a9f6ddfa54f7c6b7fe.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": "ef4f2d7ddf98871c47f562335382e6c477fd4cc8f16f90d482e7cf62ad1fc6e6", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Ik wacht aan de andere kant.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De weg gaat linksaf na de brug.\n(Natural clear speech.) [3.8 seconds duration] Loop verder... tot aan het kleine houten hek.\n(Natural clear speech.) [3.2 seconds duration] Ik wacht aan de andere kant.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "53475491a59e31bea8134e682ec04e277f3db733bee404f91690ac90db0080a4", "generation_seconds_batch": 14.36836458207108, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.537813425064087, "target_z": 1.9062498546405813, "ce": 6.181467056274414, "genuineness": 4.193545341491699, "identity": 0.650090217590332, "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_NEWS_reg/ef4f2d7ddf98871c47f562335382e6c477fd4cc8f16f90d482e7cf62ad1fc6e6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "ef7d8b8abfb5259da612e8001a13ed1c5d812f895412cfb4d306c3b56c6eb80b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6225fb4f3b672d76340eab1db155b959a9cbb184f98ed790ef7efa6e9d078864", "generation_seconds_batch": 18.725731721962802, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9695897698402405, "target_z": 0.8749999864781935, "ce": 6.165581703186035, "genuineness": 3.8420867919921875, "identity": 0.24732552468776703, "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_NEWS_reg/ef7d8b8abfb5259da612e8001a13ed1c5d812f895412cfb4d306c3b56c6eb80b.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "ef90f25ea662296bdb14e449c46c91e13d7fe4cecb368354ff9bf01c4b1101b7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "c3492944ed347df34d3c965bea67650ae981da3482681f2843ed37b260f6622b", "generation_seconds_batch": 18.862212941981852, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2709205150604248, "target_z": 1.4218749509834518, "ce": 5.927301406860352, "genuineness": 4.093128204345703, "identity": 0.6728909015655518, "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_NEWS_reg/ef90f25ea662296bdb14e449c46c91e13d7fe4cecb368354ff9bf01c4b1101b7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "ef9e72b4cd9c8edba63d0c592e3d809639f215efd8d32f13d1b87520ba0c1442", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ebf585ef56275e5162e541275fe1cf2ac8261741adfbaa1f8aa8c24788ef6f16", "generation_seconds_batch": 21.62259983702097, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9308472871780396, "target_z": 0.8046875667639191, "ce": 6.308549880981445, "genuineness": 4.65797233581543, "identity": 0.7991161942481995, "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_NEWS_reg/ef9e72b4cd9c8edba63d0c592e3d809639f215efd8d32f13d1b87520ba0c1442.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "f035ce9dc88d7536d0af49c64ea2fb979d8fa1317cc6eb5a0b82d193d289d729", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "9913e4e1309822d44fb5acb5b1f06cb240518b909c9fccd97fb37df487e5411d", "generation_seconds_batch": 21.301742951967753, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.098731517791748, "target_z": 1.1093749712661614, "ce": 6.260893821716309, "genuineness": 4.557555198669434, "identity": 0.5748320817947388, "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_NEWS_reg/f035ce9dc88d7536d0af49c64ea2fb979d8fa1317cc6eb5a0b82d193d289d729.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "f0780b28c23c95efcc101a3c9c115cabe34020517848c0fa4cee1a20cc5f2262", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d98e583c97f4903aeefd2eebf399742ca200af6b6d892ad5051d4a9f5c024f84", "generation_seconds_batch": 21.393760774983093, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6435068249702454, "target_z": 0.28320309309698805, "ce": 6.475345611572266, "genuineness": 4.532451152801514, "identity": 0.4816057085990906, "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_NEWS_reg/f0780b28c23c95efcc101a3c9c115cabe34020517848c0fa4cee1a20cc5f2262.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "f17cf7887865247f231aaaa68c2a9e5d542e35502e3c4c62424015300252765a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] De deur staat nog open.\n(Natural clear speech.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Natural clear speech.) [3.2 seconds duration] Zeg maar wanneer je klaar bent.", "seed": 1702, "scale": 1.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "2a75d0a4cf642096752c2e119095a7b09b5b1324d5cd220753fcaa0bedd6d0eb", "generation_seconds_batch": 20.330752862035297, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.460328459739685, "target_z": 1.7656250152120323, "ce": 6.157639026641846, "genuineness": 4.1998209953308105, "identity": 0.7248919010162354, "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_NEWS_reg/f17cf7887865247f231aaaa68c2a9e5d542e35502e3c4c62424015300252765a.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": "f1cb2be7163627eecef8a670cc343c6394827180b0341624cffa193e8fd594fd", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] La colocaremos después de comer.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Alguien ha movido la pequeña silla azul.\n(Natural clear speech.) [3.8 seconds duration] Estaba aquí... junto a la estantería alta.\n(Natural clear speech.) [3.2 seconds duration] La colocaremos después de comer.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "e687fa61d5362910de2f75b08b83988efa8e953a2ae07c7a05561a1c75cfd4a9", "generation_seconds_batch": 17.38785416900646, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8821914196014404, "target_z": 2.531249814075162, "ce": 6.1894097328186035, "genuineness": 4.055472373962402, "identity": 0.39094430208206177, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/f1cb2be7163627eecef8a670cc343c6394827180b0341624cffa193e8fd594fd.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "8f9b512f57e9d1c5e9e595800b2d575941eb7137e7bdfe72309f1e3faa75f599", "generation_seconds_batch": 14.116779235075228, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5520314574241638, "target_z": 0.1171875464812095, "ce": 6.268836498260498, "genuineness": 4.519899368286133, "identity": 0.5605521202087402, "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_NEWS_reg/f21bc0adf2df6aeda4aa5bc68657c60b57559ffed3b647118e970ebfd0ba6ad8.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": "f22a04f1aa54f543b086ed52fb8a98092fa35cadd4bbde529c129d5fed8d5e77", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Llévatelas cuando te vayas.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Las llaves están en la mesa de la cocina.\n(Natural clear speech.) [3.8 seconds duration] Miré en el cajón... y luego junto a la ventana.\n(Natural clear speech.) [3.2 seconds duration] Llévatelas cuando te vayas.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "21b3a92b607513fc5b4991043f4e07aa55d036d94f1ffdb498b625bb5109866a", "generation_seconds_batch": 14.463430935982615, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5808608531951904, "target_z": 1.9843751740932574, "ce": 6.2767791748046875, "genuineness": 4.068024158477783, "identity": 0.5293447971343994, "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_NEWS_reg/f22a04f1aa54f543b086ed52fb8a98092fa35cadd4bbde529c129d5fed8d5e77.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "f26f12b636cd9a21f23b2fc54f5c1a44b1f36b1580ad0e646f7eda4b6bcbaac5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "285050b2f2efbb0fba8d2c1f4fad96b9d56e7d7a65c2a52bba4e22459aa02135", "generation_seconds_batch": 19.005299902986735, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9682860374450684, "target_z": 2.6875000202827097, "ce": 6.24500846862793, "genuineness": 4.18099308013916, "identity": 0.6694102883338928, "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_NEWS_reg/f26f12b636cd9a21f23b2fc54f5c1a44b1f36b1580ad0e646f7eda4b6bcbaac5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "f2acfb3da30933b9f4410a27c9102e5a9946a794632e0e0e64fa45263d99957c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ef0f0dcee7780686eb5a87b71b35e485eab2f8bd23d0ff8ab594c21b6e290f74", "generation_seconds_batch": 19.248854364966974, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6671828031539917, "target_z": 0.3261718240819478, "ce": 6.308549880981445, "genuineness": 4.3881025314331055, "identity": 0.047723181545734406, "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_NEWS_reg/f2acfb3da30933b9f4410a27c9102e5a9946a794632e0e0e64fa45263d99957c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "f3241ba780365e5beb88c727f0ee9962782e6556ccd933ec1fddf2ee249c2bb9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0ed9db50b0af157ff11bc174b19f728a955bbf63c5d02493755f706ed092822a", "generation_seconds_batch": 19.948915104032494, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8253815174102783, "target_z": 0.613281312960911, "ce": 6.117925643920898, "genuineness": 4.281409740447998, "identity": 0.7571502923965454, "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_NEWS_reg/f3241ba780365e5beb88c727f0ee9962782e6556ccd933ec1fddf2ee249c2bb9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "f332cc76883edcbb671d9335d59741b15531f38008390de387376bab3b151cf0", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2f3e3716879244e5c626131939e6a4ca74c83aa0f4e54596ed332a25275be9a9", "generation_seconds_batch": 14.588781449012458, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0320082902908325, "target_z": 0.988281245351879, "ce": 6.086154937744141, "genuineness": 4.413206577301025, "identity": 0.7192027568817139, "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_NEWS_reg/f332cc76883edcbb671d9335d59741b15531f38008390de387376bab3b151cf0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "f33d044e803b13a4c0488e2a412caec0cc3a037a7d24e573b916acc0fe52d418", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "211aede5c9249ded606e1e08b678b8228402949fbf4c7d9716af94a0bccc29a8", "generation_seconds_batch": 21.259091073996387, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.8576669096946716, "target_z": 0.6718749780270646, "ce": 6.284721851348877, "genuineness": 4.055472373962402, "identity": 0.2738688588142395, "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_NEWS_reg/f33d044e803b13a4c0488e2a412caec0cc3a037a7d24e573b916acc0fe52d418.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": "f37c4448d88ea021d23bd01e1fd5e43e3542ce4592f08c06462ad16157cb8ef7", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ik schreef het nummer op de envelop.\n(Natural clear speech.) [3.8 seconds duration] Lees het goed... en controleer de laatste twee cijfers.\n(Natural clear speech.) [3.2 seconds duration] Bel me wanneer je even tijd hebt.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "5414121854c2b5bdabd3cbb2b9872cafc9b18a0e490e96de4d0d36611b496b50", "generation_seconds_batch": 15.226651718025096, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 2.4159774780273438, "target_z": 3.500000054087226, "ce": 6.125868320465088, "genuineness": 3.8797430992126465, "identity": 0.5751311182975769, "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_NEWS_reg/f37c4448d88ea021d23bd01e1fd5e43e3542ce4592f08c06462ad16157cb8ef7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "f396252cf52bd37d5e4bb20fc7210dec87626cb5326eff4f7c0eab110ecf16fd", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "b379af49a2c46f23f7ed97ad5940f5acd9e5eeef7f1b6ce8f6fdc64befb4330e", "generation_seconds_batch": 15.756918297032826, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3914527893066406, "target_z": 1.6406248935157746, "ce": 6.09409761428833, "genuineness": 4.325342178344727, "identity": 0.6031237840652466, "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_NEWS_reg/f396252cf52bd37d5e4bb20fc7210dec87626cb5326eff4f7c0eab110ecf16fd.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": "f3a886982102e960c6a1f399590315612b4662757494bdbc7bb173d3c6e6a144", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ton message est arrivé après mon départ.\n(Natural clear speech.) [3.8 seconds duration] Je l’ai relu... puis j’ai rangé mon téléphone.\n(Natural clear speech.) [3.2 seconds duration] Dis-moi quand nous pourrons parler.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 127, "seconds_generated": 10.16, "mp3_sha256": "70bbfed81cce043cce02a143bf9616d9f1b958b55b7a86bc3e825c45055da3f3", "generation_seconds_batch": 13.844444451970048, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1589977741241455, "target_z": 1.2187501588812253, "ce": 6.213237762451172, "genuineness": 4.268857479095459, "identity": 0.6837189197540283, "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": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/f3a886982102e960c6a1f399590315612b4662757494bdbc7bb173d3c6e6a144.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "f42016432d7c2cb762019bf4b3896b0f9ed9c5fc6fab8fe46a4c3e876371feee", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "13cdf07dd81d181f30a6b382050d0124005c0011bc583bdcda67e12ab31cae84", "generation_seconds_batch": 19.379194129025564, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5636417865753174, "target_z": 1.953124873233065, "ce": 6.197352409362793, "genuineness": 4.168440818786621, "identity": 0.6765488386154175, "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_NEWS_reg/f42016432d7c2cb762019bf4b3896b0f9ed9c5fc6fab8fe46a4c3e876371feee.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": "f42db8e407a84aba4eddb442fa4b263496dba339a379ab350192599e41e62f49", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I wrote the number on the envelope.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Call me when you have a moment.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] I wrote the number on the envelope.\n(Natural clear speech.) [3.8 seconds duration] Read it carefully... and check the last two digits.\n(Natural clear speech.) [3.2 seconds duration] Call me when you have a moment.", "seed": 1701, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "7b69ff51600892195989a74a06ea007cee647d8bd68de18d0fe58adbe002028b", "generation_seconds_batch": 14.856012354022823, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9911134243011475, "target_z": 0.9140625380300805, "ce": 6.403861999511719, "genuineness": 4.030367851257324, "identity": 0.18253979086875916, "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_NEWS_reg/f42db8e407a84aba4eddb442fa4b263496dba339a379ab350192599e41e62f49.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "f44c59449a7556ce20b6e417f7f22dbb72964cf8b65c021bb0e0ebd52d49fdb9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "0b0f0519f057afaf9bebd256ca0e448cb49ee9f2d0cdfaa27b9840b7b07c571f", "generation_seconds_batch": 15.017885050037876, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.5746312141418457, "target_z": 0.15820307957518162, "ce": 6.268836498260498, "genuineness": 4.017816066741943, "identity": 0.36928921937942505, "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_NEWS_reg/f44c59449a7556ce20b6e417f7f22dbb72964cf8b65c021bb0e0ebd52d49fdb9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "f49437c0ca60a73056c64fa79c299b020d740bb2231df3db09295f490af95778", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "06a97b562f6169d711a1c95a3f84e394d51e1f5841e80caf47d917f8e33b3046", "generation_seconds_batch": 15.728273383923806, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5894701480865479, "target_z": 1.9999998918255488, "ce": 5.879645347595215, "genuineness": 3.9425034523010254, "identity": 0.6660329103469849, "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_NEWS_reg/f49437c0ca60a73056c64fa79c299b020d740bb2231df3db09295f490af95778.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "f4da27e6d57810cecadbcaa1108100a8a6e90290a74f296dd0585ef003afceb6", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "bad77dfccb9b7755260b5dbce3490766efcedc2e4ffb7a1ed64d57aa7631795f", "generation_seconds_batch": 19.28012532601133, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.6152985095977783, "target_z": 2.0468749104180324, "ce": 6.372091293334961, "genuineness": 4.293961524963379, "identity": 0.348146915435791, "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_NEWS_reg/f4da27e6d57810cecadbcaa1108100a8a6e90290a74f296dd0585ef003afceb6.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "f523146c0e2134e7229930b0bd9307bd0b3af27fe635e014373246d08bbae7f7", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "2d9f18524832094af1b0d0810bd6472e56757005cf78bc0f91bc3c029acb6892", "generation_seconds_batch": 19.198327517020516, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.451719045639038, "target_z": 1.7500000811308385, "ce": 6.086154937744141, "genuineness": 4.231201171875, "identity": 0.6704074144363403, "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_NEWS_reg/f523146c0e2134e7229930b0bd9307bd0b3af27fe635e014373246d08bbae7f7.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-0", "key": "f52e1f6bd856b9014994e0ab526811ba68b6f76ed2cfcd4e896ccc4c72bd417b", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] De deur staat nog open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ik heb je jas op de stoel gelegd.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "596876f2548472efee201149086bf7ada4392b5bbe969a6690094fa53e1bd090", "generation_seconds_batch": 19.184261474991217, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2063496112823486, "target_z": 1.3046874045022423, "ce": 6.149696350097656, "genuineness": 4.243753433227539, "identity": 0.7748903632164001, "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_NEWS_reg/f52e1f6bd856b9014994e0ab526811ba68b6f76ed2cfcd4e896ccc4c72bd417b.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": "f5817e6c03e307d4701b94905345c252300c88ebd9a03037d452b670644ec008", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Your message arrived after I had left.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Let me know when we can talk.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Your message arrived after I had left.\n(Natural clear speech.) [3.8 seconds duration] I read it twice... before putting the phone away.\n(Natural clear speech.) [3.2 seconds duration] Let me know when we can talk.", "seed": 1702, "scale": 0.0, "cfg": 4.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "c5bffef88a0abb85f6a5bcd55c0781a3837c7fa6b42f913067482d15ce1adad3", "generation_seconds_batch": 13.917710524052382, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3096630573272705, "target_z": 1.4921874788721774, "ce": 6.284721851348877, "genuineness": 3.7918784618377686, "identity": 0.3013911545276642, "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_NEWS_reg/f5817e6c03e307d4701b94905345c252300c88ebd9a03037d452b670644ec008.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "f6b116f05e16ee682387aa0025925a236d8c19a0b526e48fbfe64b3c6e2496a1", "bucket": "vn_S_NEWS_reg", "phase": "wording", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "adf51e5f330af79386c6fabc32221de0110e840e259fb155788878633cba8e0b", "generation_seconds_batch": 15.833559063030407, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.468937873840332, "target_z": 1.7812499492932259, "ce": 6.133810997009277, "genuineness": 4.036643981933594, "identity": 0.6557657122612, "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_NEWS_reg/f6b116f05e16ee682387aa0025925a236d8c19a0b526e48fbfe64b3c6e2496a1.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "f6baf5a426ab0fb44530f5381fa057f2f1919e504a97144f3a4697d5265afacf", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "446cbe06e5769590f93110701d8a98c3c880eef1a809bbe8eac3643b9a206e59", "generation_seconds_batch": 21.27678560791537, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.098731517791748, "target_z": 1.1093749712661614, "ce": 6.332377910614014, "genuineness": 4.18099308013916, "identity": 0.4447365999221802, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/f6baf5a426ab0fb44530f5381fa057f2f1919e504a97144f3a4697d5265afacf.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "f6e857a61dfb7053daa5439cb8e8b422173faf2fd21868358bfd86efd15b4fc0", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d80e6ed3906569ce889ab5453821e527a6789ed2abfd489e0e47f2872077a7e9", "generation_seconds_batch": 19.198327517020516, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4861568212509155, "target_z": 1.812500033804516, "ce": 6.197352409362793, "genuineness": 4.557555198669434, "identity": 0.6890418529510498, "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_NEWS_reg/f6e857a61dfb7053daa5439cb8e8b422173faf2fd21868358bfd86efd15b4fc0.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-0", "key": "f7484b3da633fbcb8fd1ae7a57357f13fd6864bb6b5fba9f55cac67334982ec2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La porte est encore ouverte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] J’ai posé ton manteau sur la chaise.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "d6020fb76ce13e3aa673dd6eef39055295b69d7a79024bf98fc1ae8e3d949aed", "generation_seconds_batch": 21.77064372599125, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.047074794769287, "target_z": 1.0156249340811938, "ce": 6.0146708488464355, "genuineness": 4.469690799713135, "identity": 0.7025541067123413, "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_NEWS_reg/f7484b3da633fbcb8fd1ae7a57357f13fd6864bb6b5fba9f55cac67334982ec2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "f78f1601bcb38890af8a786a06bc60a7c74360787aa4f69cb5c5cda8b314e656", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Natural clear speech.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Natural clear speech.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "seed": 1702, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "de57f67f04ad2a68514a5c61a3e4f4a3103a7fc208e0bdfef61ed257d04292c4", "generation_seconds_batch": 21.049558890983462, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8994104862213135, "target_z": 2.5625001149353546, "ce": 6.395919322967529, "genuineness": 4.143336772918701, "identity": 0.5019153952598572, "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_NEWS_reg/f78f1601bcb38890af8a786a06bc60a7c74360787aa4f69cb5c5cda8b314e656.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "5da2e8ebb2ffb12e7cc320736ee947afa8bdf107711f860b93fbccd7265a5d47", "generation_seconds_batch": 14.338822501944378, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.2063496112823486, "target_z": 1.3046874045022423, "ce": 5.998785495758057, "genuineness": 4.3881025314331055, "identity": 0.69199538230896, "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_NEWS_reg/f7a65885d0e107603bdce9d92b5fc3b68a68524db7b0b06f79b6538794e8bf5c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "f7aad6ee37392672dc71ec89ea5937b9922ccd9cb5a3772ca0c767a44de824ac", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "fb251576beed708cd62c5a2abf9afaf98689987a0c742f44655e1ad598eb82d0", "generation_seconds_batch": 14.904808830004185, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7026968002319336, "target_z": 0.390624974646613, "ce": 6.316492557525635, "genuineness": 4.344170093536377, "identity": 0.27301162481307983, "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_NEWS_reg/f7aad6ee37392672dc71ec89ea5937b9922ccd9cb5a3772ca0c767a44de824ac.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_NEWS_reg", "phase": "heldout", "candidate": "standard", "prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": "dda052e21196448fae2aa0a972a326122cf918dc80acd03fb5d3491f46babfa0", "generation_seconds_batch": 14.950263729086146, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1934354305267334, "target_z": 1.2812498952060003, "ce": 6.292664527893066, "genuineness": 4.432034492492676, "identity": 0.4058670997619629, "hypothesis": "Die Straße führt hinter der Brücke nach links. Geh weiter bis bis zu dem kleinen Holztor. Ich warte auf der anderen Seite.", "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_NEWS_reg/f86feddc7af334e5c3d6691e715ece3168f8e7618bd3b63ca7d0171f412fdf9d.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "f8c0f65773f98f184ce53cc4354884b1462d137bbe706d92024a5d50ebd1fec9", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ddf10469751ffdc61548410207db8b7cf9825d9b58d07418a690781b008a4e50", "generation_seconds_batch": 19.057467386941426, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7272214889526367, "target_z": 2.250000135218064, "ce": 6.09409761428833, "genuineness": 4.042920112609863, "identity": 0.5715988874435425, "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_NEWS_reg/f8c0f65773f98f184ce53cc4354884b1462d137bbe706d92024a5d50ebd1fec9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "f9374bd2b8ebdf9bf362e930a46b1b21f146faa54b99b420f36f9b0e7dafcd7f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "49e078f0cc176fa12299b315063e28da547cfba3042a0f5ac5dda500fd9d6367", "generation_seconds_batch": 19.4446020249743, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9157807230949402, "target_z": 0.7773437698601532, "ce": 6.181467056274414, "genuineness": 4.595211505889893, "identity": 0.6048603653907776, "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_NEWS_reg/f9374bd2b8ebdf9bf362e930a46b1b21f146faa54b99b420f36f9b0e7dafcd7f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "f984079fd1536b695ad2fbb183ff752c0ce966af570d831a00c6a20d71a89361", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "dc7497a9fd92616b9211e15e80682b420eae6b6ad55fae89977c1eaf24afa08d", "generation_seconds_batch": 21.08110368193593, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.9394567012786865, "target_z": 0.8203125008451129, "ce": 6.173524379730225, "genuineness": 3.9927117824554443, "identity": 0.19084268808364868, "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_NEWS_reg/f984079fd1536b695ad2fbb183ff752c0ce966af570d831a00c6a20d71a89361.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "f98ad0e30f21b809e2afbf9aa7f816cc1de1cef500b7364cba94b966a29381ca", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6e914f23949c4e99d33eaf00c7d3afcb99ef7f82a934714cca4eee9eea143efc", "generation_seconds_batch": 20.16473301802762, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.6327450275421143, "target_z": 0.2636718714082702, "ce": 6.2767791748046875, "genuineness": 4.65797233581543, "identity": 0.6716888546943665, "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_NEWS_reg/f98ad0e30f21b809e2afbf9aa7f816cc1de1cef500b7364cba94b966a29381ca.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "f9ead4bb5c02a693d1e1adae8d46c491896b6232218b81c03b5c8dda3080d4b5", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "0856b544abf3d39f0c571a4ac631f37fa29aa946806efcdf3c9e0561320f5639", "generation_seconds_batch": 20.064826954971068, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7100024223327637, "target_z": 2.2187498343578715, "ce": 5.871703147888184, "genuineness": 3.835810661315918, "identity": 0.7560797333717346, "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_NEWS_reg/f9ead4bb5c02a693d1e1adae8d46c491896b6232218b81c03b5c8dda3080d4b5.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "fb09592694c890cf3b6acae93d73aa8453f774c411a31923c1020c995a627ca2", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "ef81490c4b9bde4dd21096a23b8287db7467152c3b1e5018a69f7b5dedd4fe2b", "generation_seconds_batch": 18.735959362005815, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4431095123291016, "target_z": 1.7343749307007421, "ce": 6.205295085906982, "genuineness": 4.18099308013916, "identity": 0.6898357272148132, "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_NEWS_reg/fb09592694c890cf3b6acae93d73aa8453f774c411a31923c1020c995a627ca2.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "fba588079840e56a642ad2503e4f84ce0fe135ddde512f8483c18f10937ef893", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "73d1c36e42d6ac9cd321f0b1a8465b1c6d57c5a6e1b28e8e32217a5df5857ae0", "generation_seconds_batch": 18.732180706923828, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.9596765041351318, "target_z": 2.6718748698526134, "ce": 6.229123115539551, "genuineness": 4.3127899169921875, "identity": 0.6025941967964172, "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_NEWS_reg/fba588079840e56a642ad2503e4f84ce0fe135ddde512f8483c18f10937ef893.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "fc1bbf6a5c44fc3208645371770cc8ebf86103e983b0e2355fa63c4aefb6a3d8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 1.5, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "71ef9a983085bb4eb7e6a88b770196da0d90615caeb1fc3a69ab8f514c46b91a", "generation_seconds_batch": 20.006567763979547, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3053584098815918, "target_z": 1.4843751200060318, "ce": 6.292664527893066, "genuineness": 4.281409740447998, "identity": 0.378670871257782, "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_NEWS_reg/fc1bbf6a5c44fc3208645371770cc8ebf86103e983b0e2355fa63c4aefb6a3d8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-1", "key": "fcb8ede6150bd9030e5d9faeacc94fed90a0269ab14edbfa3966d2791b38739e", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ich habe die kleine Schachtel gefunden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sie stand neben dem alten Fenster.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "63746c96bbd07e68c62b394200243977fde4a8ba0da170e9cced3a8cb50e3c4f", "generation_seconds_batch": 20.338860796997324, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.606689214706421, "target_z": 2.031250192685741, "ce": 6.395919322967529, "genuineness": 3.973883628845215, "identity": 0.20044322311878204, "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_NEWS_reg/fcb8ede6150bd9030e5d9faeacc94fed90a0269ab14edbfa3966d2791b38739e.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "fd14be84acd82d5ffc9630a439179e842de85902936dab27368f2a580ec78c38", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "6c01a18c8c7264e9baf0ea0d23804d08c2fb597317adfa7771de62af1486b3f0", "generation_seconds_batch": 18.732180706923828, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5119850635528564, "target_z": 1.8593748360480973, "ce": 6.205295085906982, "genuineness": 4.331617832183838, "identity": 0.6297406554222107, "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_NEWS_reg/fd14be84acd82d5ffc9630a439179e842de85902936dab27368f2a580ec78c38.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-nl-1", "key": "fdd7cd179e201d601d82b1f54d8c7687d09aa84ae2a68b0c599b35b4cdf1943c", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Ik heb het kleine doosje gevonden.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Het stond naast het oude raam.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "48eee089a348b4f3b6d2dcbb86dc0e28b6d43d96f9d6249c4474ab4bb0b52529", "generation_seconds_batch": 18.812548773013987, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5464229583740234, "target_z": 1.9218750050706774, "ce": 6.213237762451172, "genuineness": 4.281409740447998, "identity": 0.5370880365371704, "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_NEWS_reg/fdd7cd179e201d601d82b1f54d8c7687d09aa84ae2a68b0c599b35b4cdf1943c.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-1", "key": "fdffa7752a47e006895b4c2a28c9845f76fba9eb72e7f517a341c15df4814bba", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] I found the little box.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] It was beside the old window.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": 127, "seconds_generated": 10.16, "mp3_sha256": "8e07d27b596d29b5f7b131b131691a09d21f53bb11464625f16cb9f7f9ab4135", "generation_seconds_batch": 18.934860236011446, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.090122103691101, "target_z": 1.0937500371849675, "ce": 6.507116317749023, "genuineness": 4.444586753845215, "identity": 0.43686825037002563, "hypothesis": "I found the little box. It was beside the old window. We can look inside together.", "wer": 0.0, "duration_error_s": 0.03999999999999915, "model_predictions_not_human_ratings": true }, "audio_url": "https://huggingface.co/datasets/laion/Humaneness-Voice-Large-1.5-Intense-Bucket-LoRAs-Demo-Audio/resolve/main/manual-comparison-v1/audio/vn_S_NEWS_reg/fdffa7752a47e006895b4c2a28c9845f76fba9eb72e7f517a341c15df4814bba.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "fe0702a2852055b7086d380463468c9a08c64d616cb643d380459aaa9845f33a", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "60415f64b83ad9e5e9c60be1571dd93069f9b359594bceaec3d1a83966e2afa8", "generation_seconds_batch": 19.565134981065057, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.8477537631988525, "target_z": 2.468750077750387, "ce": 5.7684478759765625, "genuineness": 3.9801597595214844, "identity": 0.5675307512283325, "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_NEWS_reg/fe0702a2852055b7086d380463468c9a08c64d616cb643d380459aaa9845f33a.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "fe2c66cecf252bf7350f84ef5d7b2366518db508cc454936e941f14a09d6cb2f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] Die Tür ist noch offen.\n(Natural clear speech.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Natural clear speech.) [3.2 seconds duration] Sag mir, wenn du bereit bist.", "seed": 1701, "scale": 0.5, "cfg": 1.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "b81716275e7bf0601a3fedd17f4e395689bed394fee92bc351fd58c3650c27fc", "generation_seconds_batch": 20.593537491047755, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.0944268703460693, "target_z": 1.1015626124000157, "ce": 6.324435234069824, "genuineness": 4.582659721374512, "identity": 0.6237474679946899, "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_NEWS_reg/fe2c66cecf252bf7350f84ef5d7b2366518db508cc454936e941f14a09d6cb2f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-de-0", "key": "fe87a9273fa52d85c24ab2a8760577cb1483e507a472b86de3a651440ce7444f", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] Die Tür ist noch offen.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Ich habe deinen Mantel auf den Stuhl gelegt.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "12b67ea763349a23c65530bc0772f228adc64667f3b97cc416d748db38ce02e5", "generation_seconds_batch": 19.683334450935945, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.1589977741241455, "target_z": 1.2187501588812253, "ce": 6.324435234069824, "genuineness": 4.406930446624756, "identity": 0.4001964330673218, "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_NEWS_reg/fe87a9273fa52d85c24ab2a8760577cb1483e507a472b86de3a651440ce7444f.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-en-0", "key": "feb367eafe615164a69292454e0f6486ac4f11de5c0a781f907d4a72eb47d6a8", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] The door is still open.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] I put your coat on the chair.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "98f1613bf5b54e1056d028d32b183a03214aa92f0605305caab3349cff8b46ab", "generation_seconds_batch": 21.08110368193593, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 0.7134586572647095, "target_z": 0.4101563045097821, "ce": 6.2767791748046875, "genuineness": 4.055472373962402, "identity": 0.42466795444488525, "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_NEWS_reg/feb367eafe615164a69292454e0f6486ac4f11de5c0a781f907d4a72eb47d6a8.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-1", "key": "feb5b27f98428ce150ad207da788e668d35f8d4d027061da1739f6f205329382", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] He encontrado la cajita.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Estaba junto a la ventana vieja.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "aa2536adc7cb35a2ecea304ee9fa0a1f0e9f11073f8e39b2c3a051f25bcaaf8c", "generation_seconds_batch": 19.023266934091225, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3397960662841797, "target_z": 1.546874856330807, "ce": 6.165581703186035, "genuineness": 4.086852550506592, "identity": 0.5616722106933594, "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_NEWS_reg/feb5b27f98428ce150ad207da788e668d35f8d4d027061da1739f6f205329382.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-es-0", "key": "fef865f697da92e90b6a58d6693b543d14b8f6038c831e21a277da3ed2e7fac1", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La puerta sigue abierta.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] He dejado tu abrigo sobre la silla.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "30136817eb35e9e630dbedcfabd1558ca60f5490a994f936b4d6b367f14fcbe6", "generation_seconds_batch": 20.497667895047925, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.4000623226165771, "target_z": 1.6562500439458707, "ce": 5.95112943649292, "genuineness": 4.174716949462891, "identity": 0.7342321872711182, "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_NEWS_reg/fef865f697da92e90b6a58d6693b543d14b8f6038c831e21a277da3ed2e7fac1.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": "ff200cbaa706406cad30b38fd95d4ccaa96a5168d0ae4082d6284cda68c28ca9", "bucket": "vn_S_NEWS_reg", "phase": "heldout", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [3.2 seconds duration] Te esperaré al otro lado.", "neutral_prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects.\nSCRIPT:\n(Natural clear speech.) [3.2 seconds duration] La carretera gira a la izquierda tras el puente.\n(Natural clear speech.) [3.8 seconds duration] Sigue... hasta la pequeña puerta de madera.\n(Natural clear speech.) [3.2 seconds duration] Te esperaré al otro lado.", "seed": 1702, "scale": 0.0, "cfg": 5.0, "ended": true, "frames_generated": 128, "seconds_generated": 10.24, "mp3_sha256": "9810fee85fe58cab57d256d5a16d0490c36696dbe944a50f71894fb7cf53f0ef", "generation_seconds_batch": 14.552804628037848, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7358307838439941, "target_z": 2.2656248529503555, "ce": 6.221180438995361, "genuineness": 3.998987913131714, "identity": 0.6348021626472473, "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_NEWS_reg/ff200cbaa706406cad30b38fd95d4ccaa96a5168d0ae4082d6284cda68c28ca9.mp3" }, { "id": "vn_S_NEWS_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_NEWS_reg-fr-1", "key": "fff9aa1e2bb0de3965cbf36cb577b56dcbbe027752406e22749cc1cd28f187ac", "bucket": "vn_S_NEWS_reg", "phase": "grid", "candidate": "manual-59e65fa10d8f", "prompt": "GENERAL: One adult speaker, clean high-quality recording, no music or sound effects. Deliver with a highly polished, almost academic neutrality, prioritizing impeccable articulation.\nSCRIPT:\n(Start with a crisp, clear articulation that establishes immediate factual distance.) [3.2 seconds duration] J’ai trouvé la petite boîte.\n(Ensure the pace remains consistently regulated, avoiding any natural speech ebb and flow.) [3.8 seconds duration] Elle était près de la vieille fenêtre.\n(Preserve this highly controlled, precise delivery right up to the very last sound.) [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": "1b02eaefba9c691cefd973bda3da48a3e182fb78138116242948f57d3eeaefa4", "generation_seconds_batch": 19.293313191039488, "original_audio_reused": false, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.5292041301727295, "target_z": 1.8906251369082898, "ce": 5.959072113037109, "genuineness": 4.18726921081543, "identity": 0.6525803804397583, "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_NEWS_reg/fff9aa1e2bb0de3965cbf36cb577b56dcbbe027752406e22749cc1cd28f187ac.mp3" }, { "key": "original-vn_S_NEWS_reg-de-0-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "2cf2a0cb2e668311f76d446260e8c7a248ae9dc18f89d52e1f9ab976da91e55b", "task_id": "vn_S_NEWS_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.3053584098815918, "target_z": 1.4843751200060318, "ce": 6.086154937744141, "genuineness": 4.193545341491699, "identity": 0.45039841532707214, "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_NEWS_reg/de/030da9e51f4e7b40e1e48cfc.mp3" }, { "key": "original-vn_S_NEWS_reg-de-0-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "da509a2a63aa0b2cc6f393fa0a022177aa9e190af10ee137148ce6d230c5f4cb", "task_id": "vn_S_NEWS_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.3699291944503784, "target_z": 1.601562450138339, "ce": 6.23706579208374, "genuineness": 4.469690799713135, "identity": 0.6453597545623779, "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_NEWS_reg/de/19ad2ffa4cdb4ff9b8af9b21.mp3" }, { "key": "original-vn_S_NEWS_reg-de-0-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "3a0874cef9efc5f4aee89e735463139173733a80f0c374b2674e6703d0dd3632", "task_id": "vn_S_NEWS_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.3053584098815918, "target_z": 1.4843751200060318, "ce": 6.09409761428833, "genuineness": 4.281409740447998, "identity": 0.49733594059944153, "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_NEWS_reg/de/01c108d4718ddfea084e6f49.mp3" }, { "key": "original-vn_S_NEWS_reg-de-0-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "3c808499170aff48758a27fd9470af2ac4ba3d2c8cbbe34ecdcb193e21bfc8e7", "task_id": "vn_S_NEWS_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.4000623226165771, "target_z": 1.6562500439458707, "ce": 6.300607204437256, "genuineness": 4.362998008728027, "identity": 0.5863048434257507, "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_NEWS_reg/de/a3206922a1b398624f544b87.mp3" }, { "key": "original-vn_S_NEWS_reg-de-0-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "e5ef8fbbec30cc486c54e89468137edd0e2ed9713466933a9ec2d4a0c406b7a9", "task_id": "vn_S_NEWS_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.5122127532958984, "target_z": 0.04492192887594738, "ce": 6.2767791748046875, "genuineness": 4.645420074462891, "identity": 0.4876289963722229, "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_NEWS_reg/de/0a8d328fa29a3ee67d715b4b.mp3" }, { "key": "original-vn_S_NEWS_reg-de-0-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "69c21cbe30b8a7c814fb891fb455993f998bc5aa21731cb69d1c1473dd08891c", "task_id": "vn_S_NEWS_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.666106641292572, "target_z": 0.3242187343654114, "ce": 6.284721851348877, "genuineness": 4.833701133728027, "identity": 0.599459171295166, "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_NEWS_reg/de/7ae62af9ae3b5dcbecdfc19d.mp3" }, { "key": "original-vn_S_NEWS_reg-de-0-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "d6fbdc59b485270697a5006e3f795703d6bda6ba777ec97e9d60a7ad0599db2d", "task_id": "vn_S_NEWS_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.1460835933685303, "target_z": 1.1953126495849833, "ce": 6.205295085906982, "genuineness": 4.532451152801514, "identity": 0.509781002998352, "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_NEWS_reg/de/7eb09a318b966108ef9174e0.mp3" }, { "key": "original-vn_S_NEWS_reg-de-0-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "876c4762ac16c69d470928921c534ce28d940303665efb5acdaff467a2966d40", "task_id": "vn_S_NEWS_reg-de-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.20204496383667, "target_z": 1.2968750456360967, "ce": 6.181467056274414, "genuineness": 4.344170093536377, "identity": 0.4991191327571869, "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_NEWS_reg/de/8d6ca481a380eb513881fd7e.mp3" }, { "key": "original-vn_S_NEWS_reg-de-1-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "90b3fd522c69a1ad1be9b865ff8b07c5cb3ecf8245f68041e932cf76d68dec35", "task_id": "vn_S_NEWS_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7186119556427002, "target_z": 2.234374984787968, "ce": 6.507116317749023, "genuineness": 4.2249250411987305, "identity": 0.5250630378723145, "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_NEWS_reg/de/c8716aabdf6e010d446d3801.mp3" }, { "key": "original-vn_S_NEWS_reg-de-1-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "e324a8f9a988fd12d0e7f26769421b1cf42f2e7584325885c92a0a437a77269c", "task_id": "vn_S_NEWS_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7186119556427002, "target_z": 2.234374984787968, "ce": 6.3641486167907715, "genuineness": 4.162164688110352, "identity": 0.5147007703781128, "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_NEWS_reg/de/38c6c9d5188ab3364c1714e8.mp3" }, { "key": "original-vn_S_NEWS_reg-de-1-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "eb8ad62b82de64363012ac09a1e48a87a3bc1701c6a4fea8e008d82d8318b55a", "task_id": "vn_S_NEWS_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.520594596862793, "target_z": 1.8749999864781937, "ce": 6.435632705688477, "genuineness": 4.319066047668457, "identity": 0.580348551273346, "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_NEWS_reg/de/3fd6780d44e1040e9235bd53.mp3" }, { "key": "original-vn_S_NEWS_reg-de-1-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "c9789dba8b1a6a2a26f9c3d68df98a56beef965a17f28082dd2b6e51b8d3d2dd", "task_id": "vn_S_NEWS_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7100024223327637, "target_z": 2.2187498343578715, "ce": 6.435632705688477, "genuineness": 4.256305694580078, "identity": 0.5950590372085571, "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_NEWS_reg/de/e54dd93b9e88ac6228894326.mp3" }, { "key": "original-vn_S_NEWS_reg-de-1-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "f1556eba1817deb0cec12b7e7d6d224072f8b3622d0ce3e2f87827d1ade5ade7", "task_id": "vn_S_NEWS_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.7186119556427002, "target_z": 2.234374984787968, "ce": 6.435632705688477, "genuineness": 4.118232727050781, "identity": 0.5358768701553345, "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_NEWS_reg/de/64ed5026712f93f1ba923e6c.mp3" }, { "key": "original-vn_S_NEWS_reg-de-1-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "050543f40b0d99417a65188720160b3868512801b1fbcb529896ea6888bf9cb6", "task_id": "vn_S_NEWS_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3828433752059937, "target_z": 1.6249999594345808, "ce": 6.38003396987915, "genuineness": 4.306513786315918, "identity": 0.34850215911865234, "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_NEWS_reg/de/eae600e14d9d54bc4e52c95d.mp3" }, { "key": "original-vn_S_NEWS_reg-de-1-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "f24a1bb168ce4d4e7933a39f805c390d13a18b824c448de17cfcb8311412c7ea", "task_id": "vn_S_NEWS_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.417281150817871, "target_z": 1.6874999121082583, "ce": 6.395919322967529, "genuineness": 4.268857479095459, "identity": 0.5828118324279785, "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_NEWS_reg/de/d44c5f3d854cbf58287ff84d.mp3" }, { "key": "original-vn_S_NEWS_reg-de-1-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "977b2cdd6b4f805d63fded37ce0aec54f1523de14fd4e4a6d19c8a402258a18d", "task_id": "vn_S_NEWS_reg-de-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.) [3.2 seconds duration] Wir können gemeinsam hineinschauen.", "original_audio_reused": true, "scoring_contract": "singleton-EARS-ASR-Orange128-preMP3-v1", "metrics": { "valid_waveform": true, "target_raw": 1.3225772380828857, "target_z": 1.5156249881684194, "ce": 6.395919322967529, "genuineness": 4.457139015197754, "identity": 0.48502814769744873, "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_NEWS_reg/de/5a689e9f4c863c09f4e8b5ac.mp3" }, { "key": "original-vn_S_NEWS_reg-en-0-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "919086cc067dede30af12454dbbf9200ce25138c40da68e8ca07b3dadfdd9791", "task_id": "vn_S_NEWS_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.9222378134727478, "target_z": 0.7890625245082741, "ce": 6.300607204437256, "genuineness": 4.231201171875, "identity": 0.20994900166988373, "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_NEWS_reg/en/de5591bfc013de43003b1217.mp3" }, { "key": "original-vn_S_NEWS_reg-en-0-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "6a1f405348fc17fea63e520d6b6c5f2fe8582991371ae0b368c5528fc9b93817", "task_id": "vn_S_NEWS_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.5337363481521606, "target_z": 0.08398437225338308, "ce": 6.3641486167907715, "genuineness": 4.2374773025512695, "identity": 0.2487489879131317, "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_NEWS_reg/en/49a89b67a9f64aa00e6abad6.mp3" }, { "key": "original-vn_S_NEWS_reg-en-0-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "d38364da52cc11e4ae6e3f78d8631152b83d32b54dbdddea5cc262a3fa8f6c6d", "task_id": "vn_S_NEWS_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.6704113483428955, "target_z": 0.3320312014060082, "ce": 6.181467056274414, "genuineness": 4.20609712600708, "identity": 0.0838979035615921, "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_NEWS_reg/en/0249d37a4775c33f6f0a05a1.mp3" }, { "key": "original-vn_S_NEWS_reg-en-0-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "2bb8da24b9feb4d97bf778382a255a34a5b25aab64fe64810732173465aef643", "task_id": "vn_S_NEWS_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.5337363481521606, "target_z": 0.08398437225338308, "ce": 6.308549880981445, "genuineness": 4.193545341491699, "identity": 0.3956281244754791, "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_NEWS_reg/en/e22aadf8aaf07f80124a4495.mp3" }, { "key": "original-vn_S_NEWS_reg-en-0-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "5424a9978a59efd5952a8f9c599190febdde4567c9a4ec8b6b5e1cf86741d5ec", "task_id": "vn_S_NEWS_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8619716167449951, "target_z": 0.6796874450676615, "ce": 6.340320587158203, "genuineness": 4.030367851257324, "identity": 0.14638671278953552, "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_NEWS_reg/en/58d1ee71bb42b22d6ed0367a.mp3" }, { "key": "original-vn_S_NEWS_reg-en-0-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "9a23e125f89df6d9cce83dadf9eb173b735c709f7da7f9a8a5f38aa08887e346", "task_id": "vn_S_NEWS_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8834952712059021, "target_z": 0.7187499966195484, "ce": 6.109982967376709, "genuineness": 3.8169827461242676, "identity": 0.2659093141555786, "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_NEWS_reg/en/08300e628423b56d80c47a8f.mp3" }, { "key": "original-vn_S_NEWS_reg-en-0-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "16d71691f82f5f868e0b39877bb1747842bb91a03da232c71bcf022b809ce5d2", "task_id": "vn_S_NEWS_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8404480218887329, "target_z": 0.6406250016902258, "ce": 6.252951145172119, "genuineness": 4.055472373962402, "identity": 0.22631633281707764, "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_NEWS_reg/en/ca13a4e8ea786f2b76de2d1c.mp3" }, { "key": "original-vn_S_NEWS_reg-en-0-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "4a298cfc5983190039f0a31747a4b392307d50a020f3ef1387199847628f8f59", "task_id": "vn_S_NEWS_reg-en-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.9179331064224243, "target_z": 0.7812500574676772, "ce": 6.165581703186035, "genuineness": 3.6663577556610107, "identity": 0.159726083278656, "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_NEWS_reg/en/f550a1c75cc21b3311193b95.mp3" }, { "key": "original-vn_S_NEWS_reg-en-1-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "c1208915b08c0185c39baafa98b7cd5833fc21a7fd8bab9281efe68e6f961636", "task_id": "vn_S_NEWS_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.40567076206207275, "target_z": -0.1484375228180483, "ce": 6.3641486167907715, "genuineness": 4.65797233581543, "identity": 0.35420531034469604, "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_NEWS_reg/en/5b4aa79049251de0e4129a3d.mp3" }, { "key": "original-vn_S_NEWS_reg-en-1-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "49b6df091efae85df9eabdfe65cbafd9e26764a8838e5eb01d64d3f55e6b19eb", "task_id": "vn_S_NEWS_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.47347021102905273, "target_z": -0.025390599012778323, "ce": 6.443574905395508, "genuineness": 4.607763767242432, "identity": 0.3333529829978943, "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_NEWS_reg/en/f1ec790e7b5b409958bab0fd.mp3" }, { "key": "original-vn_S_NEWS_reg-en-1-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "7f7f6664a801a26165ec0c18cab88cb6a3c8894a020886c977af63b9765de822", "task_id": "vn_S_NEWS_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.7758771181106567, "target_z": 0.5234374552090163, "ce": 6.459460258483887, "genuineness": 4.425758361816406, "identity": 0.40446120500564575, "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_NEWS_reg/en/16c31035d566f2e1393bf667.mp3" }, { "key": "original-vn_S_NEWS_reg-en-1-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "5bddc39f883b8af5772cebe23fbff3357dc539de0dc49a20ab7626a44b6c493f", "task_id": "vn_S_NEWS_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.45947983860969543, "target_z": -0.05078125211278225, "ce": 6.213237762451172, "genuineness": 4.607763767242432, "identity": 0.6424936652183533, "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_NEWS_reg/en/0922b5b3785610e457101f01.mp3" }, { "key": "original-vn_S_NEWS_reg-en-1-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "28f6afedd44fa1ab19a1ccd9d4c8335bdbd9a0c73f5a026afcab48acdce133fb", "task_id": "vn_S_NEWS_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.5197460055351257, "target_z": 0.05859377324060475, "ce": 6.395919322967529, "genuineness": 4.545003414154053, "identity": 0.6998772621154785, "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_NEWS_reg/en/3688f2a953f46bab9aa1e4d0.mp3" }, { "key": "original-vn_S_NEWS_reg-en-1-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "42de09e4fe6a6653a12c66d612d14305770478a7c776d24340963959de0d7fe7", "task_id": "vn_S_NEWS_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.5864692330360413, "target_z": 0.1796874991548871, "ce": 6.308549880981445, "genuineness": 4.325342178344727, "identity": 0.2666972875595093, "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_NEWS_reg/en/c02eb4f72151da9dec43dadc.mp3" }, { "key": "original-vn_S_NEWS_reg-en-1-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "b51042db591316df6c4ae1f2bd925143f1a8a3a3d55590c5a54c529b7c5774ea", "task_id": "vn_S_NEWS_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.5746312141418457, "target_z": 0.15820307957518162, "ce": 6.229123115539551, "genuineness": 4.369274139404297, "identity": 0.46515533328056335, "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_NEWS_reg/en/8b86b6aff57b145a1fe31b2a.mp3" }, { "key": "original-vn_S_NEWS_reg-en-1-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "09488c4730f8f61f34f21c7f99a0461cc6d89e511b3cf61e4b99acaa903c29c1", "task_id": "vn_S_NEWS_reg-en-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.43365147709846497, "target_z": -0.09765627070526606, "ce": 6.395919322967529, "genuineness": 4.507347106933594, "identity": 0.24346047639846802, "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_NEWS_reg/en/c05cccb7abce5bb735da860b.mp3" }, { "key": "original-vn_S_NEWS_reg-es-0-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "72d5614b97217a6d49c6230e4bae7c5e2cfb57f5db277e2f9b7bcd2bcb1a9276", "task_id": "vn_S_NEWS_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.210654377937317, "target_z": 1.3124999797172905, "ce": 6.141753673553467, "genuineness": 4.080576419830322, "identity": 0.47487926483154297, "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_NEWS_reg/es/addb39757dd4ef11956b6183.mp3" }, { "key": "original-vn_S_NEWS_reg-es-0-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "5408ebd37f7325028187d6aeb7878a43c25d66a3d86d51bbfda3b591b9bb148c", "task_id": "vn_S_NEWS_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.8305349349975586, "target_z": 2.4375002095879994, "ce": 5.792276382446289, "genuineness": 3.785602569580078, "identity": 0.614814281463623, "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_NEWS_reg/es/4b0ca35308822b8f919e80e5.mp3" }, { "key": "original-vn_S_NEWS_reg-es-0-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "c3cb5104d2caf0a96f9ddd40047d45965354e901555fc7af5d3af0a5832ac1ee", "task_id": "vn_S_NEWS_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.3914527893066406, "target_z": 1.6406248935157746, "ce": 5.982900142669678, "genuineness": 3.9425034523010254, "identity": 0.6691681742668152, "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_NEWS_reg/es/aa962529e23a2a36abbeec6a.mp3" }, { "key": "original-vn_S_NEWS_reg-es-0-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "1c57c8c78577182b5566b89dfd965cabb0551529ebb0f73620ff8f3dc06a9c4e", "task_id": "vn_S_NEWS_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.4818520545959473, "target_z": 1.8046874585894679, "ce": 6.125868320465088, "genuineness": 3.9550557136535645, "identity": 0.4698787331581116, "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_NEWS_reg/es/4f0e093b5443a2d7c0f0d9c4.mp3" }, { "key": "original-vn_S_NEWS_reg-es-0-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "882ed32dcf2935fe329ef1949674e6d079b6c06a66a775d8fcedb85569551d9e", "task_id": "vn_S_NEWS_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.2192637920379639, "target_z": 1.3281249137984843, "ce": 6.038498878479004, "genuineness": 4.20609712600708, "identity": 0.541612446308136, "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_NEWS_reg/es/31de7aa707d3a13ffec9125e.mp3" }, { "key": "original-vn_S_NEWS_reg-es-0-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1702, "scale": 0.5, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "9d2d9a85889ef1305452fc8c6001f2f5930af98bca10198bd7d37f23d0cc456d", "task_id": "vn_S_NEWS_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.2278733253479004, "target_z": 1.3437500642285805, "ce": 5.764476776123047, "genuineness": 4.124508857727051, "identity": 0.4567703604698181, "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_NEWS_reg/es/e1e548f211c82b4778ec2423.mp3" }, { "key": "original-vn_S_NEWS_reg-es-0-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "cb10bc9545f79d949982cbfc49d7e68d300cb599c87da31b87eaeeaee36fea54", "task_id": "vn_S_NEWS_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.4086717367172241, "target_z": 1.6718749780270645, "ce": 5.990842819213867, "genuineness": 4.149612903594971, "identity": 0.7224714756011963, "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_NEWS_reg/es/faf7e3a5607790e7ef712c1b.mp3" }, { "key": "original-vn_S_NEWS_reg-es-0-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "7bc4a768d44108c0068076b010d293696f50217a415c38b8fe335195c125a3c7", "task_id": "vn_S_NEWS_reg-es-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.081512689590454, "target_z": 1.0781251031037737, "ce": 6.022613525390625, "genuineness": 4.15588903427124, "identity": 0.5230532884597778, "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_NEWS_reg/es/4d5b41dfeeb87bc9b51f0712.mp3" }, { "key": "original-vn_S_NEWS_reg-es-1-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "30b202bf92c0a7bbce2761c032c8780c722357686c90465c35c15b414c1fa64c", "task_id": "vn_S_NEWS_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.205295085906982, "genuineness": 4.432034492492676, "identity": 0.3252450227737427, "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_NEWS_reg/es/58bb9a9cfe7583df0ecdcd81.mp3" }, { "key": "original-vn_S_NEWS_reg-es-1-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "b1b07f1f2aa9947dd56afcb630eb106337ae6ba30214f2c99c82c25d0993bc60", "task_id": "vn_S_NEWS_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.6841740608215332, "target_z": 2.171874815765388, "ce": 6.117925643920898, "genuineness": 4.494794845581055, "identity": 0.6542537808418274, "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_NEWS_reg/es/d83435eb5150bf51d5ed7d15.mp3" }, { "key": "original-vn_S_NEWS_reg-es-1-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "bfd6e0221e793b12085950ee3e6494b066750777d8d086d38634bb99c4d18cc6", "task_id": "vn_S_NEWS_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.3914527893066406, "target_z": 1.6406248935157746, "ce": 6.125868320465088, "genuineness": 4.3881025314331055, "identity": 0.7086373567581177, "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_NEWS_reg/es/fd0ca40769433fec6cb02f44.mp3" }, { "key": "original-vn_S_NEWS_reg-es-1-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "f3a55b9d76259511923104d7228750686eba44b9c464e96e2f363477022b4bed", "task_id": "vn_S_NEWS_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.1589977741241455, "target_z": 1.2187501588812253, "ce": 6.221180438995361, "genuineness": 4.570107460021973, "identity": 0.5597378611564636, "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_NEWS_reg/es/8f44dc314df07c9dffde8fb0.mp3" }, { "key": "original-vn_S_NEWS_reg-es-1-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "ad806f722705a7257a506ffd949422a9d0628a116ac6981238e5560ce8652e3e", "task_id": "vn_S_NEWS_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.9631326794624329, "target_z": 0.8632812318300727, "ce": 6.078212261199951, "genuineness": 4.65797233581543, "identity": 0.4998655915260315, "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_NEWS_reg/es/e06ac182405eadb6ff270311.mp3" }, { "key": "original-vn_S_NEWS_reg-es-1-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "c4c008629227fbad49e71643d2e82b43ed5d27786bf2d0d27c2f639433a892ed", "task_id": "vn_S_NEWS_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8017054796218872, "target_z": 0.5703124738015001, "ce": 6.173524379730225, "genuineness": 4.871357440948486, "identity": 0.6764742136001587, "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_NEWS_reg/es/8a892de87b5ce0229f683b90.mp3" }, { "key": "original-vn_S_NEWS_reg-es-1-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "4db5667dce41deed7f8d9997abfd49c3f67e860223dacb8548517898326cd23a", "task_id": "vn_S_NEWS_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.7930960655212402, "target_z": 0.5546875397203063, "ce": 5.967014789581299, "genuineness": 4.507347106933594, "identity": 0.6098936796188354, "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_NEWS_reg/es/d0467a9c514991311009a10f.mp3" }, { "key": "original-vn_S_NEWS_reg-es-1-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "b3a2235f579815cf5edad20344db26a23578ce26ac9d8b44dd2925075301ee7a", "task_id": "vn_S_NEWS_reg-es-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.973894476890564, "target_z": 0.8828124535187905, "ce": 6.109982967376709, "genuineness": 4.645420074462891, "identity": 0.7949259877204895, "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_NEWS_reg/es/dd3df1a20af13a44ef65dad5.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-0-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "99b027533210fcd388a694b15631379306eb1c8cadd620a5754627049c82c7d2", "task_id": "vn_S_NEWS_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.6047643423080444, "target_z": 0.21289067338271353, "ce": 6.316492557525635, "genuineness": 4.6956281661987305, "identity": 0.7890511751174927, "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_NEWS_reg/fr/afb4734366f3292ad6c700f7.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-0-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "c60e3c0f01f1ed5f9aa6e14e7bb6acfe8f268498efe98d950deea89e99ac24d0", "task_id": "vn_S_NEWS_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.1374740600585938, "target_z": 1.1796874991548871, "ce": 5.990842819213867, "genuineness": 4.356721878051758, "identity": 0.7879409790039062, "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_NEWS_reg/fr/b8c0e45390924e1452faf358.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-0-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "4f36453d8114f8ecc7ead72d560e15a42aea0e1492b8baccf6ed5602603112e4", "task_id": "vn_S_NEWS_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8426003456115723, "target_z": 0.6445311811232987, "ce": 6.260893821716309, "genuineness": 4.406930446624756, "identity": 0.5879223942756653, "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_NEWS_reg/fr/84775969ca5951eb3eac160a.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-0-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "86f385f4328b334a2d5023c615f331f48dc26a3ff01bf307b1cfa6289cb952b2", "task_id": "vn_S_NEWS_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.7629629373550415, "target_z": 0.4999999459127744, "ce": 6.046441555023193, "genuineness": 4.632867813110352, "identity": 0.7559158205986023, "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_NEWS_reg/fr/fe54974f6a1ad6075eb878e6.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-0-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "51d17c7a55d543111302751b6ea172424d46b937f6eb64968f36a8479c057397", "task_id": "vn_S_NEWS_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.2881394624710083, "target_z": 1.4531250354947418, "ce": 6.24500846862793, "genuineness": 4.293961524963379, "identity": 0.5708250999450684, "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_NEWS_reg/fr/1902ffc8609552fa19c2a482.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-0-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "d5fad114c151a0ea362f3aee8c91ae5e1e921277501a7695b91849add12b6a99", "task_id": "vn_S_NEWS_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.9695897698402405, "target_z": 0.8749999864781935, "ce": 6.173524379730225, "genuineness": 4.394378662109375, "identity": 0.5226330161094666, "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_NEWS_reg/fr/6cd6f3dc0b204762001df88b.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-0-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "74408de2a0625939eb39b1dd2af46bd0caefbab467c05ddf095d5d65571ff550", "task_id": "vn_S_NEWS_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.115950345993042, "target_z": 1.140624839428549, "ce": 6.1894097328186035, "genuineness": 4.256305694580078, "identity": 0.8220773935317993, "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_NEWS_reg/fr/bca0cdc3cb57b9e310ec4323.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-0-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "c445374ec52b705ac4cc0f599c72a9dbd7b171fdb54bfd8b7e870f9d72c8ed0e", "task_id": "vn_S_NEWS_reg-fr-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.3010536432266235, "target_z": 1.4765625447909838, "ce": 6.070269584655762, "genuineness": 4.400654315948486, "identity": 0.6925160884857178, "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_NEWS_reg/fr/714c9a90c69d0751253dc81b.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-1-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "1688e9ccd915a8b43d07ac5174edb0e1e473931d0ea5ba08d89e10fa5f372f7a", "task_id": "vn_S_NEWS_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8770381808280945, "target_z": 0.7070312419714274, "ce": 6.165581703186035, "genuineness": 4.632867813110352, "identity": 0.7124803066253662, "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_NEWS_reg/fr/542162a11864500f69287549.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-1-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "faf77962e4f0c681bc65735eae94fb1e022e78f3eef000573b88762f6d5011b4", "task_id": "vn_S_NEWS_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8060102462768555, "target_z": 0.5781250490165482, "ce": 6.23706579208374, "genuineness": 4.645420074462891, "identity": 0.6275187730789185, "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_NEWS_reg/fr/1bce31d8e86abc0fc93ee28f.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-1-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "19ae284a2d7835a31ae10550dafe9b62dba5e711407e84ab187644ba69583e35", "task_id": "vn_S_NEWS_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.0104846954345703, "target_z": 0.9492188019744433, "ce": 6.205295085906982, "genuineness": 4.344170093536377, "identity": 0.7398273944854736, "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_NEWS_reg/fr/63aaffeeb2ef75665bf90444.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-1-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "25d42848d384bfc844a7a8625b5dd44831623000bdf0d1078c42e4d6d4168d85", "task_id": "vn_S_NEWS_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.266615867614746, "target_z": 1.414062592117306, "ce": 6.165581703186035, "genuineness": 4.519899368286133, "identity": 0.5854787230491638, "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_NEWS_reg/fr/7a87e756b3400eb3683808e1.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-1-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "9d7535c9a7cf28479ca38573924e81555e7d0aea7c56d0e90e4a730a14c255fb", "task_id": "vn_S_NEWS_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.1848260164260864, "target_z": 1.2656249611248065, "ce": 6.1894097328186035, "genuineness": 4.2751336097717285, "identity": 0.6356253623962402, "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_NEWS_reg/fr/602e4ba05e5b0814da430862.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-1-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "f301ffb9244d83ed93ea6f68efc71bf348b9266a146a07ba2c293f0ddfd8cc89", "task_id": "vn_S_NEWS_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8921047449111938, "target_z": 0.7343750388751934, "ce": 6.173524379730225, "genuineness": 4.507347106933594, "identity": 0.796212911605835, "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_NEWS_reg/fr/6044ef27603f759555ed118f.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-1-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "d435f7f1ffeea983d115c682938afc04a5f3caf2d6e8e23773dc963ddf0ca75e", "task_id": "vn_S_NEWS_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.0944268703460693, "target_z": 1.1015626124000157, "ce": 6.062326908111572, "genuineness": 4.268857479095459, "identity": 0.779069185256958, "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_NEWS_reg/fr/0a06873376dda9c2abf5a399.mp3" }, { "key": "original-vn_S_NEWS_reg-fr-1-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "ee102f0a54a3435154a7c5d25aa01fa440e00fcc006f6b0a303bcc837283390d", "task_id": "vn_S_NEWS_reg-fr-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8189244270324707, "target_z": 0.6015625583127902, "ce": 6.268836498260498, "genuineness": 4.7458367347717285, "identity": 0.7720047235488892, "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_NEWS_reg/fr/2a212cfd3cccdc50d56e5b34.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-0-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "cbc147ed27d547807a1250b5df92729f12057a4087bfd30ad35097edfd0a30b0", "task_id": "vn_S_NEWS_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.1288645267486572, "target_z": 1.164062348724791, "ce": 5.967014789581299, "genuineness": 4.1998209953308105, "identity": 0.47048628330230713, "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_NEWS_reg/nl/c8dfb3add9e48e77db93a252.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-0-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "cd0b6188dfeba256fe23e996282c87af9b2706fcc12fa8c5ffd6628914f215cd", "task_id": "vn_S_NEWS_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8856476545333862, "target_z": 0.7226562842270724, "ce": 6.133810997009277, "genuineness": 4.381826400756836, "identity": 0.6198074817657471, "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_NEWS_reg/nl/a4ab3590b19a7623009779ab.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-0-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word." ], "seconds": 10.2, "frames": 128, "seeds": [ 1701, 1702 ], "scales": [ 0.0, 0.5, 0.75, 1.0 ], "cfg": 1.0, "reference": true, "case_type": "Neutral wording, target emotion/style varies only through acoustic performance", "seed": 1701, "scale": 0.75, "frames_generated": 127, "seconds_generated": 10.16, "ended": true, "contract_sha256": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "74a94fe2d193a209a22713e0958c2620454113dbf57999db68c864e4704fd376", "task_id": "vn_S_NEWS_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.7500487565994263, "target_z": 0.47656243661653247, "ce": 5.990842819213867, "genuineness": 4.337893962860107, "identity": 0.474634051322937, "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_NEWS_reg/nl/d97f6a7671a8ff889325e9c5.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-0-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "e8b94fe1c5c16be06bfc681688cd6c66fe89d15e3be5ae26a01af99fe8aa0c64", "task_id": "vn_S_NEWS_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.6983920931816101, "target_z": 0.3828125076060161, "ce": 5.935244083404541, "genuineness": 4.036643981933594, "identity": 0.5576226711273193, "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_NEWS_reg/nl/c9d4d5ec0970dd162d381ac8.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-0-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "5ffd4c6c1cb5c792e6b15c10443d59a7127e70dbe1defc789a209f2edfa174b7", "task_id": "vn_S_NEWS_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.210654377937317, "target_z": 1.3124999797172905, "ce": 5.816103935241699, "genuineness": 4.607763767242432, "identity": 0.5544260144233704, "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_NEWS_reg/nl/28bba193045820bbbd6a36fd.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-0-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "99dfff2afed09d93a746c7c5d5502522fbc925259f027fd052e2224cbf08ca68", "task_id": "vn_S_NEWS_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.7651153206825256, "target_z": 0.5039062335202984, "ce": 6.157639026641846, "genuineness": 4.394378662109375, "identity": 0.6240730285644531, "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_NEWS_reg/nl/940c63d4ec96488354a8f1a6.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-0-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "ff9e9a446311ecb0a56d31b42d757d532313afa49fa2b35bac816964137e91ce", "task_id": "vn_S_NEWS_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.810314953327179, "target_z": 0.5859375160571451, "ce": 5.95112943649292, "genuineness": 4.394378662109375, "identity": 0.7504888772964478, "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_NEWS_reg/nl/4ec3c8f0eb9d4a3ad8be1b74.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-0-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "b9b5dac1d304018b06b0bd2dc75006eaed7bd44c7fae0eacc5150b23b4c5c22d", "task_id": "vn_S_NEWS_reg-nl-0", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.8877999782562256, "target_z": 0.7265624636601453, "ce": 6.109982967376709, "genuineness": 4.356721878051758, "identity": 0.7587337493896484, "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_NEWS_reg/nl/ef43313f8e9e4ec8e42cec66.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-1-seed1701-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "fd7c6738f962d5b3115911e0e937c0d0639c85ceae98fa7728bebb9422a7b741", "task_id": "vn_S_NEWS_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.0729031562805176, "target_z": 1.0624999526736776, "ce": 6.197352409362793, "genuineness": 4.425758361816406, "identity": 0.6232348680496216, "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_NEWS_reg/nl/b8802c038ec8d68e88138df2.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-1-seed1701-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "6b4687128100eb60e885ec2a00ba8294f4c76325da9ce09216b584a0f0caf055", "task_id": "vn_S_NEWS_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.025551199913025, "target_z": 0.9765624907037581, "ce": 6.141753673553467, "genuineness": 4.457139015197754, "identity": 0.5986316204071045, "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_NEWS_reg/nl/b0078efe9107957b93ed241b.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-1-seed1701-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "a7f3c8cbe299e26ae8e4f51659472462e33dc090ca2f564fffbc807593430321", "task_id": "vn_S_NEWS_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.7651153206825256, "target_z": 0.5039062335202984, "ce": 6.165581703186035, "genuineness": 4.306513786315918, "identity": 0.7784304618835449, "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_NEWS_reg/nl/666bf316505a852e83cc5625.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-1-seed1701-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "fa641a9c490cfd92b279b76032b44a3c92acb34b52aa017e96bd31465f770228", "task_id": "vn_S_NEWS_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.016941785812378, "target_z": 0.9609375566225643, "ce": 6.2767791748046875, "genuineness": 4.369274139404297, "identity": 0.7824355363845825, "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_NEWS_reg/nl/8c29f95f5629e7b11702c4b9.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-1-seed1702-scale0.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "0e6f9084a0462bca034412f863cce83903002b1230a0f601367a6fdb8eb38dc3", "task_id": "vn_S_NEWS_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.6583456993103027, "target_z": 2.124999797172904, "ce": 6.078212261199951, "genuineness": 4.2249250411987305, "identity": 0.6929519176483154, "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_NEWS_reg/nl/4a42493d2a341428226facbc.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-1-seed1702-scale0.5", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "2be879ac4b97541b736ac2caf63ae63a041a6819a6d68313aff3a87c98b1dd31", "task_id": "vn_S_NEWS_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.9846563339233398, "target_z": 0.9023437833819595, "ce": 6.23706579208374, "genuineness": 4.362998008728027, "identity": 0.6611884832382202, "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_NEWS_reg/nl/db53f6262cb140c74ac3441b.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-1-seed1702-scale0.75", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "de12d469236ea518a420cb3a18170006cade902067cb4dfc4b9a7baa0e77b4de", "task_id": "vn_S_NEWS_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 1.4345000982284546, "target_z": 1.7187499966195483, "ce": 6.0146708488464355, "genuineness": 4.124508857727051, "identity": 0.6108627319335938, "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_NEWS_reg/nl/2243c3bbbc8d12f908a2e35e.mp3" }, { "key": "original-vn_S_NEWS_reg-nl-1-seed1702-scale1.0", "bucket": "vn_S_NEWS_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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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 Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.", "Strong Newsreader 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": "ee0ffc384b74920011e7eba05cf7076318375aea270b40aea44090a1e2d8fc72", "mp3_sha256": "762440099a9b34399fa8e9f86783f29fa36106364771aaff2e92f6d888258710", "task_id": "vn_S_NEWS_reg-nl-1", "phase": "standard", "candidate": "standard", "neutral_prompt": "GENERAL: One adult speaker in a clean recording. Strong Newsreader Style. Make the requested delivery clearly audible, not merely implied by the words. Preserve every word.\nSCRIPT:\n(Strong Newsreader 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 Newsreader 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 Newsreader 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": 0.9825039505958557, "target_z": 0.8984374957744355, "ce": 6.054384231567383, "genuineness": 4.369274139404297, "identity": 0.6055932641029358, "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_NEWS_reg/nl/a5fb91327ec1d2f840ecae6e.mp3" } ], "source_standard_audio_unchanged": true }