Paradis-TTS / provenance.json
IvannP's picture
Record successful PyPI 0.1.3 publication; mirror verified distributions and pinned registry
df0c94f verified
Raw History Blame Contribute Delete
26.2 kB
{
"release": "v0.1.3",
"package_version": "0.1.3",
"default_weight_kind": "ema",
"parameters": {
"decoder": 3851053,
"text": 4231828,
"total": 8082881
},
"text": {
"fingerprint": "f6046c2195f96114615f63a941af1f0706e3270d7ad6ad013363ba1100a30c12",
"source_checkpoint_sha256": "e779be24b5a70c1832c9b3946a581b031053dc50362478e702dbd07e9d77d340",
"step": 8000
},
"training": {
"status": "COMPLETE",
"step": 10000,
"ema_updates": 10000,
"ema_decay": 0.999,
"fingerprint": "7ad5a1e89d672e521a4e11fbae1e22d84a88363572d5ba13cf88a3c9ca48e2e6",
"last_checkpoint_sha256": "f0cb9a15781c3699534dbdb3206027b4b02529e337bdb5c7599e61d0180aab3a",
"validation": {
"ema": {
"mel": 0.5812129572629932,
"stft": 0.9282944535017015,
"total": 1.5095074107646946
},
"raw": {
"mel": 0.5816698493957513,
"stft": 0.9281938228607175,
"total": 1.509863672256469
}
}
},
"inference_defaults": {
"phase_lock_frames": 97,
"presence_restore": true,
"presence_max_gain_db": 2.5,
"dynamic_soften_db": 0.0,
"upper_soften_db": 1.5,
"attack_soften_db": 0.0,
"syllable_soften_db": 0.0,
"dynamics_smoothing": 0.0,
"softening_match_level": true
},
"model_license": "Apache-2.0",
"perceptual_gain_established": false,
"historical_evaluation": "The 300-sentence v0 quality scores use raw step 6500; not rerun for v0.1.3.",
"exports": {
"ema": {
"checkpoint": {
"kind": "ema",
"path": "step_10000_ema.pt",
"sha256": "5e37e1d3af4ac970003c741c21b3ad5643546d1f9d6b0d43d3b546daf5881d61",
"step": 10000
},
"parity": [
{
"errors": {
"duration": 1.9073486328125e-06,
"f0": 9.1552734375e-05,
"features": 1.0132789611816406e-06,
"noise": 2.86102294921875e-06,
"prosody": 1.7881393432617188e-06,
"waveform": 3.9517879486083984e-05,
"waveform_relative_rms": 3.265972456040478e-05
},
"samples": 30000,
"speed": 1.0,
"text": "Bonjour."
},
{
"errors": {
"duration": 9.5367431640625e-07,
"f0": 0.0001068115234375,
"features": 1.1026859283447266e-06,
"noise": 4.76837158203125e-06,
"prosody": 2.6226043701171875e-06,
"waveform": 0.00027666985988616943,
"waveform_relative_rms": 0.00010339898632497038
},
"samples": 76200,
"speed": 0.75,
"text": "Ces six saucisses sont assez salées."
},
{
"errors": {
"duration": 3.814697265625e-06,
"f0": 7.62939453125e-05,
"features": 1.0728836059570312e-06,
"noise": 3.814697265625e-06,
"prosody": 2.7418136596679688e-06,
"waveform": 1.4603137969970703e-06,
"waveform_relative_rms": 1.1573270547322356e-06
},
"samples": 66600,
"speed": 1.25,
"text": "La lune illumine le village au milieu de la nuit."
},
{
"errors": {
"duration": 1.9073486328125e-06,
"f0": 0.0001220703125,
"features": 1.1324882507324219e-06,
"noise": 5.7220458984375e-06,
"prosody": 3.0994415283203125e-06,
"waveform": 1.3932585716247559e-06,
"waveform_relative_rms": 1.3723806186955894e-06
},
"samples": 133200,
"speed": 1.0,
"text": "Le billet coûte 12,50 € et le train part à 18 h 30."
}
],
"portable_audio": {
"checks": [
{
"matched_noise_source_max_abs_error": 7.367134094238281e-05,
"postprocess_max_abs_error": 2.682209014892578e-06,
"text": "Bonjour."
},
{
"matched_noise_source_max_abs_error": 0.0001372992992401123,
"postprocess_max_abs_error": 6.6533684730529785e-06,
"text": "Ces six saucisses sont assez salées."
},
{
"matched_noise_source_max_abs_error": 0.0001935809850692749,
"postprocess_max_abs_error": 3.5297125577926636e-06,
"text": "La lune illumine le village au milieu de la nuit."
}
],
"source_rng_note": "Matched noise checks interpolation/math; production PCG64 noise differs from Torch RNG",
"status": "PORTABLE_AUDIO_PARITY_PASS"
},
"quantization": {
"decoder": {
"compute": "float32 after dequantization",
"int8_weight_elements": 3814432,
"quantized_tensors": 127
},
"prosody": {
"compute": "float32 after dequantization",
"int8_weight_elements": 1113600,
"quantized_tensors": 29
},
"text": {
"compute": "float32 after dequantization",
"int8_weight_elements": 3031552,
"quantized_tensors": 35
}
},
"onnx_sizes": {
"fp32": 32819396,
"int8": 9186652
},
"offline_checks": {
"fp32": {
"status": "OFFLINE_PORTABLE_BUNDLE_PASS",
"variant": "fp32",
"torch_installed": false,
"kokoro_installed": false,
"network_blocked": true,
"seed_repeatable": true,
"different_seed_changes_audio": true,
"speed_05_and_2_pass": true,
"long_text_chunks": 2,
"wav_exports": 3,
"pcm_sample_rate": 24000,
"standalone_postprocess_pass": true,
"settings_file_pass": true,
"editable_default_settings_pass": true,
"samples": [
{
"text": "Le TGV part à 18 h 30 et le billet coûte 12,50 €.",
"normalized_text": "Le té gé vé part à dix-huit heures trente et le billet coûte douze euros et cinquante centimes.",
"seconds": 5.7,
"elapsed_seconds": 0.5648492006585002,
"rtf": 0.09909635099271932,
"peak": 0.5805146098136902,
"rms": 0.07124831141030535,
"clipped_fraction": 0.0
},
{
"text": "Voir https://exemple.fr ou écrire à contact@exemple.fr.",
"normalized_text": "Voir exemple point èf ère ou écrire à contact arobase exemple point èf ère.",
"seconds": 5.35,
"elapsed_seconds": 0.47355119977146387,
"rtf": 0.08851424294793718,
"peak": 0.6202261447906494,
"rms": 0.06945659843026926,
"clipped_fraction": 0.0
}
]
},
"int8": {
"status": "OFFLINE_PORTABLE_BUNDLE_PASS",
"variant": "int8",
"torch_installed": false,
"kokoro_installed": false,
"network_blocked": true,
"seed_repeatable": true,
"different_seed_changes_audio": true,
"speed_05_and_2_pass": true,
"long_text_chunks": 2,
"wav_exports": 3,
"pcm_sample_rate": 24000,
"standalone_postprocess_pass": true,
"settings_file_pass": true,
"editable_default_settings_pass": true,
"samples": [
{
"text": "Le TGV part à 18 h 30 et le billet coûte 12,50 €.",
"normalized_text": "Le té gé vé part à dix-huit heures trente et le billet coûte douze euros et cinquante centimes.",
"seconds": 5.675,
"elapsed_seconds": 0.5392949003726244,
"rtf": 0.09502993839165189,
"peak": 0.6017069816589355,
"rms": 0.0706897177619429,
"clipped_fraction": 0.0
},
{
"text": "Voir https://exemple.fr ou écrire à contact@exemple.fr.",
"normalized_text": "Voir exemple point èf ère ou écrire à contact arobase exemple point èf ère.",
"seconds": 5.4,
"elapsed_seconds": 0.4996072007343173,
"rtf": 0.09251985198783653,
"peak": 0.6003544926643372,
"rms": 0.06867472747507007,
"clipped_fraction": 0.0
}
]
}
}
},
"raw": {
"checkpoint": {
"kind": "raw",
"path": "step_10000_raw.pt",
"sha256": "48f897e60e68a6efca7fdb469e8310b53949f31cc21614c46d6eb201ac2c7507",
"step": 10000
},
"parity": [
{
"errors": {
"duration": 1.9073486328125e-06,
"f0": 9.1552734375e-05,
"features": 1.0132789611816406e-06,
"noise": 2.86102294921875e-06,
"prosody": 1.7881393432617188e-06,
"waveform": 3.9964914321899414e-05,
"waveform_relative_rms": 3.322003797492771e-05
},
"samples": 30000,
"speed": 1.0,
"text": "Bonjour."
},
{
"errors": {
"duration": 9.5367431640625e-07,
"f0": 0.0001068115234375,
"features": 1.1026859283447266e-06,
"noise": 4.76837158203125e-06,
"prosody": 2.6226043701171875e-06,
"waveform": 0.0003320276737213135,
"waveform_relative_rms": 0.00012161555858409843
},
"samples": 76200,
"speed": 0.75,
"text": "Ces six saucisses sont assez salées."
},
{
"errors": {
"duration": 3.814697265625e-06,
"f0": 7.62939453125e-05,
"features": 1.0728836059570312e-06,
"noise": 3.814697265625e-06,
"prosody": 2.7418136596679688e-06,
"waveform": 1.1920928955078125e-06,
"waveform_relative_rms": 1.1487845144308893e-06
},
"samples": 66600,
"speed": 1.25,
"text": "La lune illumine le village au milieu de la nuit."
},
{
"errors": {
"duration": 1.9073486328125e-06,
"f0": 0.0001220703125,
"features": 1.1324882507324219e-06,
"noise": 5.7220458984375e-06,
"prosody": 3.0994415283203125e-06,
"waveform": 1.4426186680793762e-06,
"waveform_relative_rms": 1.2788401025664708e-06
},
"samples": 133200,
"speed": 1.0,
"text": "Le billet coûte 12,50 € et le train part à 18 h 30."
}
],
"portable_audio": {
"checks": [
{
"matched_noise_source_max_abs_error": 7.367134094238281e-05,
"postprocess_max_abs_error": 2.6673078536987305e-06,
"text": "Bonjour."
},
{
"matched_noise_source_max_abs_error": 0.0001372992992401123,
"postprocess_max_abs_error": 6.307382136583328e-06,
"text": "Ces six saucisses sont assez salées."
},
{
"matched_noise_source_max_abs_error": 0.0001935809850692749,
"postprocess_max_abs_error": 3.682449460029602e-06,
"text": "La lune illumine le village au milieu de la nuit."
}
],
"source_rng_note": "Matched noise checks interpolation/math; production PCG64 noise differs from Torch RNG",
"status": "PORTABLE_AUDIO_PARITY_PASS"
},
"quantization": {
"decoder": {
"compute": "float32 after dequantization",
"int8_weight_elements": 3814432,
"quantized_tensors": 127
},
"prosody": {
"compute": "float32 after dequantization",
"int8_weight_elements": 1113600,
"quantized_tensors": 29
},
"text": {
"compute": "float32 after dequantization",
"int8_weight_elements": 3031552,
"quantized_tensors": 35
}
},
"onnx_sizes": {
"fp32": 32819396,
"int8": 9186652
},
"offline_checks": {
"fp32": {
"status": "OFFLINE_PORTABLE_BUNDLE_PASS",
"variant": "fp32",
"torch_installed": false,
"kokoro_installed": false,
"network_blocked": true,
"seed_repeatable": true,
"different_seed_changes_audio": true,
"speed_05_and_2_pass": true,
"long_text_chunks": 2,
"wav_exports": 3,
"pcm_sample_rate": 24000,
"standalone_postprocess_pass": true,
"settings_file_pass": true,
"editable_default_settings_pass": true,
"samples": [
{
"text": "Le TGV part à 18 h 30 et le billet coûte 12,50 €.",
"normalized_text": "Le té gé vé part à dix-huit heures trente et le billet coûte douze euros et cinquante centimes.",
"seconds": 5.7,
"elapsed_seconds": 0.562054899521172,
"rtf": 0.09860612272301264,
"peak": 0.5909550189971924,
"rms": 0.07189265014241372,
"clipped_fraction": 0.0
},
{
"text": "Voir https://exemple.fr ou écrire à contact@exemple.fr.",
"normalized_text": "Voir exemple point èf ère ou écrire à contact arobase exemple point èf ère.",
"seconds": 5.35,
"elapsed_seconds": 0.47477260045707226,
"rtf": 0.08874254214150884,
"peak": 0.6295928955078125,
"rms": 0.0700346613322121,
"clipped_fraction": 0.0
}
]
},
"int8": {
"status": "OFFLINE_PORTABLE_BUNDLE_PASS",
"variant": "int8",
"torch_installed": false,
"kokoro_installed": false,
"network_blocked": true,
"seed_repeatable": true,
"different_seed_changes_audio": true,
"speed_05_and_2_pass": true,
"long_text_chunks": 2,
"wav_exports": 3,
"pcm_sample_rate": 24000,
"standalone_postprocess_pass": true,
"settings_file_pass": true,
"editable_default_settings_pass": true,
"samples": [
{
"text": "Le TGV part à 18 h 30 et le billet coûte 12,50 €.",
"normalized_text": "Le té gé vé part à dix-huit heures trente et le billet coûte douze euros et cinquante centimes.",
"seconds": 5.675,
"elapsed_seconds": 0.5372727997601032,
"rtf": 0.09467362110310189,
"peak": 0.6043785214424133,
"rms": 0.07153125129059947,
"clipped_fraction": 0.0
},
{
"text": "Voir https://exemple.fr ou écrire à contact@exemple.fr.",
"normalized_text": "Voir exemple point èf ère ou écrire à contact arobase exemple point èf ère.",
"seconds": 5.4,
"elapsed_seconds": 0.494446599856019,
"rtf": 0.09156418515852204,
"peak": 0.6135249137878418,
"rms": 0.06944268676750295,
"clipped_fraction": 0.0
}
]
}
}
}
},
"release_assets": {
"release": "v0.1.3",
"package_version": "0.1.3",
"model_repository": "IvannP/Paradis-TTS",
"model_revision": "3815937c283e9f63aa63791d3b5455b9270e426f",
"default_weight_kind": "ema",
"variants": {
"fp32": {
"url": "https://huggingface.co/IvannP/Paradis-TTS/resolve/3815937c283e9f63aa63791d3b5455b9270e426f/paradis_tts_fp32.zip",
"sha256": "95d251d3ac5baf9f6d6aefa7b657973e6cd7c8a73e8afcd45f718c5f8d2997b7",
"bytes": 30214783,
"onnx_bytes": 32819396
},
"int8": {
"url": "https://huggingface.co/IvannP/Paradis-TTS/resolve/3815937c283e9f63aa63791d3b5455b9270e426f/paradis_tts_int8.zip",
"sha256": "94cd8bed0761d09d40339b4ba6d6488015049c8961033086902399df91dd9fee",
"bytes": 8214915,
"onnx_bytes": 9186652
}
},
"alternative_weights": {
"raw": {
"fp32": {
"url": "https://huggingface.co/IvannP/Paradis-TTS/resolve/3815937c283e9f63aa63791d3b5455b9270e426f/paradis_tts_raw_fp32.zip",
"sha256": "aceab20a98281ba1e1844b25d19df1f2db2e3100d6b6c300eae98da283dce331",
"bytes": 30214371,
"onnx_bytes": 32819396
},
"int8": {
"url": "https://huggingface.co/IvannP/Paradis-TTS/resolve/3815937c283e9f63aa63791d3b5455b9270e426f/paradis_tts_raw_int8.zip",
"sha256": "c0fa3feb21dfc43545a164ed315b64ebd3e9bf7e54735a48f184143744437f06",
"bytes": 8215150,
"onnx_bytes": 9186652
}
}
}
},
"derived_from": "v0.1.2; all neural graph and checkpoint hashes unchanged",
"runtime_changes": [
"Optimized NumPy filters",
"Dynamic softening disabled by default"
],
"speed_evaluation": "evaluation/speed_20261016/report.json",
"offline_validation": {
"status": "FOUR_BUNDLES_PASS",
"no_network": true,
"no_torch_loaded": true,
"checks": [
{
"kind": "ema",
"variant": "fp32",
"status": "PASS",
"samples": [
{
"text": "Bonjour, voici la voix française de Paradis.",
"seed": 1729,
"seconds": 3.175,
"elapsed_seconds": 0.2510240999981761,
"rtf": 0.07906270866084288,
"peak": 0.6160742044448853,
"rms": 0.0746403694686982,
"clipped_fraction": 0.0
},
{
"text": "Sophie sèche ses chaussettes sur sa chaise.",
"seed": 1730,
"seconds": 2.675,
"elapsed_seconds": 0.2010124996304512,
"rtf": 0.07514485967493503,
"peak": 0.42525041103363037,
"rms": 0.06898909980174596,
"clipped_fraction": 0.0
},
{
"text": "Le TGV part à 18 h 30 et le billet coûte 12,50 €.",
"seed": 1731,
"seconds": 5.7,
"elapsed_seconds": 0.4390938999131322,
"rtf": 0.07703401752861969,
"peak": 0.5893058180809021,
"rms": 0.07127009404326864,
"clipped_fraction": 0.0
}
],
"deterministic_seed": true,
"speed_control": true,
"long_text_chunks": 2,
"postprocessing": {
"phase_lock_frames": 97,
"presence_restore": true,
"presence_max_gain_db": 2.5,
"dynamic_soften_db": 0.0,
"upper_soften_db": 1.5,
"attack_soften_db": 0.0,
"syllable_soften_db": 0.0,
"dynamics_smoothing": 0.0,
"softening_match_level": true
},
"neural_file_hashes": {
"decoder.onnx": "cf2c0ef8fe8e7ef26b6930c326bdb17bf8e8a3741e949033f326b79c0786f95a",
"prosody.onnx": "d76cc58a1df1a98f4bf67eabbe5df41761f3c6c1b1d22a82e2a215a257408274",
"text.onnx": "aa93a503dfabe8f3d476740329481bb13724118b36c9c1780df8676a4258794a",
"vocab.json": "02762226a6edbbfb28295d0617b333484bf821a3785a59f3cc4b8d1f32a6f2cf"
}
},
{
"kind": "ema",
"variant": "int8",
"status": "PASS",
"samples": [
{
"text": "Bonjour, voici la voix française de Paradis.",
"seed": 1729,
"seconds": 3.175,
"elapsed_seconds": 0.2231115996837616,
"rtf": 0.07027136997913752,
"peak": 0.6272944211959839,
"rms": 0.07406429844834687,
"clipped_fraction": 0.0
},
{
"text": "Sophie sèche ses chaussettes sur sa chaise.",
"seed": 1730,
"seconds": 2.675,
"elapsed_seconds": 0.18049420043826103,
"rtf": 0.06747446745355554,
"peak": 0.41465139389038086,
"rms": 0.06832051826131692,
"clipped_fraction": 0.0
},
{
"text": "Le TGV part à 18 h 30 et le billet coûte 12,50 €.",
"seed": 1731,
"seconds": 5.675,
"elapsed_seconds": 0.39074999932199717,
"rtf": 0.06885462543118893,
"peak": 0.5817117691040039,
"rms": 0.07070998258180439,
"clipped_fraction": 0.0
}
],
"deterministic_seed": true,
"speed_control": true,
"long_text_chunks": 2,
"postprocessing": {
"phase_lock_frames": 97,
"presence_restore": true,
"presence_max_gain_db": 2.5,
"dynamic_soften_db": 0.0,
"upper_soften_db": 1.5,
"attack_soften_db": 0.0,
"syllable_soften_db": 0.0,
"dynamics_smoothing": 0.0,
"softening_match_level": true
},
"neural_file_hashes": {
"decoder.onnx": "047cd14e5b38c90be23c2c7d68b213b085bf8fb10cfe9b2808f75c10cf110c20",
"prosody.onnx": "ac8d84cf9813dc105d3fd6080802c65a92bc1cdb680f341b13df4283cae82460",
"text.onnx": "1e063f73d4ae771eaae37972169a389eb0e4db5497d428086df50044c0d83bd2",
"vocab.json": "02762226a6edbbfb28295d0617b333484bf821a3785a59f3cc4b8d1f32a6f2cf"
}
},
{
"kind": "raw",
"variant": "fp32",
"status": "PASS",
"samples": [
{
"text": "Bonjour, voici la voix française de Paradis.",
"seed": 1729,
"seconds": 3.175,
"elapsed_seconds": 0.22527700010687113,
"rtf": 0.07095338586043186,
"peak": 0.6231405735015869,
"rms": 0.07526868851579864,
"clipped_fraction": 0.0
},
{
"text": "Sophie sèche ses chaussettes sur sa chaise.",
"seed": 1730,
"seconds": 2.675,
"elapsed_seconds": 0.18473329953849316,
"rtf": 0.06905917739756753,
"peak": 0.42551061511039734,
"rms": 0.06941511552013564,
"clipped_fraction": 0.0
},
{
"text": "Le TGV part à 18 h 30 et le billet coûte 12,50 €.",
"seed": 1731,
"seconds": 5.7,
"elapsed_seconds": 0.3975323000922799,
"rtf": 0.06974250878811929,
"peak": 0.6011703014373779,
"rms": 0.07191660610670365,
"clipped_fraction": 0.0
}
],
"deterministic_seed": true,
"speed_control": true,
"long_text_chunks": 2,
"postprocessing": {
"phase_lock_frames": 97,
"presence_restore": true,
"presence_max_gain_db": 2.5,
"dynamic_soften_db": 0.0,
"upper_soften_db": 1.5,
"attack_soften_db": 0.0,
"syllable_soften_db": 0.0,
"dynamics_smoothing": 0.0,
"softening_match_level": true
},
"neural_file_hashes": {
"decoder.onnx": "6f2a5aa5f7dd616a1c77c2597cadee1aad3adb82437f3e498d57272f1b783d7c",
"prosody.onnx": "d76cc58a1df1a98f4bf67eabbe5df41761f3c6c1b1d22a82e2a215a257408274",
"text.onnx": "aa93a503dfabe8f3d476740329481bb13724118b36c9c1780df8676a4258794a",
"vocab.json": "02762226a6edbbfb28295d0617b333484bf821a3785a59f3cc4b8d1f32a6f2cf"
}
},
{
"kind": "raw",
"variant": "int8",
"status": "PASS",
"samples": [
{
"text": "Bonjour, voici la voix française de Paradis.",
"seed": 1729,
"seconds": 3.175,
"elapsed_seconds": 0.21606780029833317,
"rtf": 0.06805285048766399,
"peak": 0.6294447779655457,
"rms": 0.07492364447748445,
"clipped_fraction": 0.0
},
{
"text": "Sophie sèche ses chaussettes sur sa chaise.",
"seed": 1730,
"seconds": 2.675,
"elapsed_seconds": 0.1777013000100851,
"rtf": 0.06643039252713462,
"peak": 0.4174092710018158,
"rms": 0.06889307180634527,
"clipped_fraction": 0.0
},
{
"text": "Le TGV part à 18 h 30 et le billet coûte 12,50 €.",
"seed": 1731,
"seconds": 5.675,
"elapsed_seconds": 0.4086148999631405,
"rtf": 0.07200262554416573,
"peak": 0.5890954732894897,
"rms": 0.07155166526023406,
"clipped_fraction": 0.0
}
],
"deterministic_seed": true,
"speed_control": true,
"long_text_chunks": 2,
"postprocessing": {
"phase_lock_frames": 97,
"presence_restore": true,
"presence_max_gain_db": 2.5,
"dynamic_soften_db": 0.0,
"upper_soften_db": 1.5,
"attack_soften_db": 0.0,
"syllable_soften_db": 0.0,
"dynamics_smoothing": 0.0,
"softening_match_level": true
},
"neural_file_hashes": {
"decoder.onnx": "e705632e8c4cfcff91eaeea6ca4d2478b3de1c80da09e5a953f50a083ee10d35",
"prosody.onnx": "ac8d84cf9813dc105d3fd6080802c65a92bc1cdb680f341b13df4283cae82460",
"text.onnx": "1e063f73d4ae771eaae37972169a389eb0e4db5497d428086df50044c0d83bd2",
"vocab.json": "02762226a6edbbfb28295d0617b333484bf821a3785a59f3cc4b8d1f32a6f2cf"
}
}
]
}
}