WideCodec — 44.1 kHz decoder
A 0.8 kbps, 44.1 kHz audio tokenizer. Decoder finetune of
neuphonic/neucodec.
| sample rate | 44,100 Hz |
| tokens/sec | 50 |
| codebooks | 1 (FSQ, levels=[4]×8) |
| bitrate | 0.8 kbps |
| hop | 882 |
Codes are identical to base NeuCodec. Only the reconstruction sample rate and fidelity differ.
Quickstart
Everything needed is in this repo — the neucodec/ package and infer_widecodec.py.
pip install torch transformers huggingface_hub local-attention einops librosa soundfile
huggingface-cli download Scicom-intl/WideCodec --local-dir WideCodec
cd WideCodec
python infer_widecodec.py --input my.wav --out-dir out # file or directory
import sys; sys.path.insert(0, "WideCodec")
import torch, librosa, soundfile as sf
from neucodec import NeuCodec
model = NeuCodec._from_pretrained(model_id="Scicom-intl/WideCodec", decoder_depth=20)
model = model.eval().cuda()
wav16, _ = librosa.load("my.wav", sr=16000, mono=True) # encoder ingests 16 kHz mono
x = torch.from_numpy(wav16).float().view(1, 1, -1).cuda()
with torch.no_grad():
codes = model.encode_code(x) # 0.8 kbps, 50 tok/s
wav44 = model.decode_code(codes) # 44.1 kHz
sf.write("recon.wav", wav44.squeeze().cpu().numpy(), model.sample_rate)
decoder_depth=20 is required. The weights are a depth-20 decoder; any other value
mismatches the architecture.
Benchmark
9,291 clips — 50 each from 188 clean ≥44.1 kHz datasets, 400+ language/dialect tags. faster-UTMOSv2 plus spectral measures against ground truth.
| codec | native SR | tokens/s | codebooks | bitrate | UTMOSv2 | mel-L1 ↓ | HF≥11k | rolloff |
|---|---|---|---|---|---|---|---|---|
| ground truth | — | — | — | — | 2.822 | — | 0.00244 | 8263 |
| nvidia nemo44k | 44.1k | ~86 | RVQ (many) | ~6–9 kbps | 2.903 | 0.379 | 0.00196 | 8372 |
| WideCodec (2026-08-16) | 44.1k | 50 | 1 | ~0.8 kbps | 2.919 | 0.571 | 0.00223 | 9011 |
| WideCodec (2026-07-21) | 44.1k | 50 | 1 | ~0.8 kbps | 2.788 | 0.571 | 0.00223 | 9011 |
| dac | 44.1k | 86 | 9 | ~8 kbps | 2.672 | 0.341 | 0.00156 | 8185 |
| snac44k | 44.1k | multi-scale | 3–4 | ~2.6 kbps | 2.340 | 0.493 | 0.00158 | 8402 |
| encodec48k | 48k | 150 | RVQ | 24 kbps | 2.042 | 0.458 | 0.00169 | 8800 |
UTMOSv2 2.919 — above ground truth (2.822), level with NVIDIA NeMo-44k (2.903), at about a tenth of its bitrate.
Bitrate = frame rate × Σ log₂(codebook size). Here: 8 · log₂4 = 16 bits/frame × 50 frames/s = 800 bps. DAC: 9 codebooks × 10 bits × 86 fps ≈ 8 kbps.
The 2026-08-16 release gained +0.131 UTMOSv2 over the previous one at unchanged content accuracy — CV22 macro CER 7.70% for both, inside the ±0.25 pt run-to-run spread. Same architecture, same losses, same 50 tokens/s and single codebook. Only the training corpus changed, to clean DNSMOS-gated ≥44.1 kHz audio.
Not a benchmark artifact: on a held-out set of 1,330 clips from 138 sources absent from training it scores 2.807 against 2.654 (+0.153, paired t = +14.2), and PESQ-WB improves too (+0.017, t = +7.5). Naturalness and fidelity moved together.
Full report: evaluation/REPORT_44k_class_codecs.md.
Files and revisions
| file | what |
|---|---|
pytorch_model.bin |
decoder weights for inference |
last.ckpt |
full Lightning checkpoint with optimizer states, for resuming training |
Two training stages. main points at the latest.
| stage | data | how to load |
|---|---|---|
| 1 — noisy mix (steps 0→1.58M) | 8 base corpora + scale44k, ~6,500 h, mostly noisy crowdsourced/podcast |
pin revision d2c612f4319c1c1e0c65a9431a7d6db1b7955ce0 |
| 2 — clean finetune (from 1.58M) | TTS-Clean44k + Clean-Podcast + clean-teacher pool + EARS/Expresso | main |
model = NeuCodec._from_pretrained(
model_id="Scicom-intl/WideCodec",
decoder_depth=20,
revision="d2c612f4319c1c1e0c65a9431a7d6db1b7955ce0", # stage 1
)
Training data
High-sample-rate human speech, no synthetic TTS. The scale-up group is SR-verified ≥ 44 kHz (probed before inclusion).
Base corpora:
| dataset | content |
|---|---|
malaysia-ai/malaysian-podcast-youtube |
Malay long-form podcasts (~2,234 h) |
malaysia-ai/singaporean-podcast-youtube |
Singaporean English podcasts (~1,255 h) |
malaysia-ai/Multilingual-TTS |
CommonVoice-sidon multilingual short clips |
malaysia-ai/malaysian-cartoons-youtube |
48 kHz cartoon dialogue |
malaysia-ai/malaysian-movie-youtube |
48 kHz Malay movie dialogue |
ylacombe/expresso |
48 kHz expressive read speech |
nytopop/expresso-conversational |
48 kHz improvised dialogue |
| EARS | 48 kHz anechoic fullband |
Scale-up: 336 further datasets (162 at 48 kHz, 174 at 44.1 kHz), mirrored as
<name>*audio.zip in
malaysia-ai/Multilingual-TTS.
The decoder is language-agnostic — the codebook is frozen — so this is for acoustic diversity.
1rsh/gujarati-f-openslr— 48000 HzAashish17405/audio-dataset— 48000 HzAashish17405/audio-dataset-300— 48000 HzAashish17405/audio-dataset-shuffled-300— 48000 HzAbdullah500/IndicTTS-Bengali— 48000 HzAbdullah500/IndicTTS_BengaliOLD— 48000 Hzadalat-ai/in22-legal— 48000 Hzahmadafaneh/common-voice-18-arabic— 48000 HzAIDC-AI/CSEMOTIONS— 48000 HzAJosh/audio-dataset— 48000 Hzalexantonov/chuvash_voice— 48000 Hzaliyzd95/common_voice_21_0_fa— 48000 Hzaliyzd95/common_voice_22_0_fa— 48000 Hzanian0707/hindi-tts-dataset— 48000 Hzatlithor/talromur3_with_prompts— 48000 Hzatlithor/talromur3_without_emotions— 48000 Hzbilguun/cv-mn-24.0— 48000 Hzbookbot/slr72_dataset— 48000 HzChingkheinganba/IndicTTS_Manipuri— 48000 Hzchuuhtetnaing/myanmar-speech-dataset-openslr-80— 48000 HzCnam-LMSSC/vibravox_enhanced_by_EBEN— 48000 HzCraneAILabs/waxal-lug-clean— 48000 Hzdatahiveai/arabic-multidialect-emotional-speech-demo— 48000 HzDatarrX/burmese-synthetic-speech-corpus— 48000 HzDDD-Cambodia/khm-asr-cultural— 48000 Hzdeepdml/igbo-dict— 48000 Hzdeepdml/igbo-dict-16khz— 48000 Hzdeepdml/igbo-dict-expansion— 48000 Hzdeepdml/igbo-dict-expansion-16khz— 48000 Hzdeepdml/openslr-32-hq-SA-languages— 48000 Hzdeepdml/openslr42-khmer-tts— 48000 Hzdeepdml/openslr65-tamil— 48000 Hzdeepdml/openslr80-burmese— 48000 Hzdoof-ferb/fpt_fosd— 48000 Hzdoof-ferb/infore1_25hours— 48000 Hzespnet/ace-kising-segments— 48000 Hzespnet/ace-opencpop-segments— 48000 HzFatimahEmadEldin/alsanaa-emirati-arabic-asr— 48000 Hzfluffypotatoes/f1-team-radio— 48000 Hzfosters/lagodny-tsmok-iury-zhygamont-output_original— 48000 Hzfosters/lagodny-tsmok-iury-zhygamont_all— 48000 Hzgauravparajuli/slr43— 48000 Hzggfox00000/stt-summre-fr-test— 48000 Hzggfox00000/stt-vibravox-fr-test— 48000 Hzhanamizuki-ai/genshin-voice-v3.3-mandarin— 48000 Hzhanamizuki-ai/genshin-voice-v3.4-mandarin— 48000 Hzhanamizuki-ai/genshin-voice-v3.5-mandarin— 48000 HzHeshamHaroon/arabic-msa-25k-saudi-male-tashkeel— 48000 HzHeshamHaroon/Dahee7— 48000 Hzhezarai/common-voice-13-fa— 48000 Hzhosein-m/french_homophone_asr— 48000 Hzhumyn-labs/Asian-High-Fidelity-ASR-Dataset— 48000 Hzhumyn-labs/LATAM-High-Fidelity-ASR— 48000 Hzhypaai/Hypa-Speech-10k— 48000 Hzhypaai/Hypa_Fleurs— 48000 Hzigidn/wuwa-voice-EN— 48000 Hzimpriyanshu-garg00/IndicVoices-R_Hindi— 48000 HzJacobLinCool/jl-speech— 48000 HzJeanKouss/ewe_bible_v2_tts— 48000 Hzjspaulsen/vctk— 48000 Hzjuanjucm/OpenHQ-SpeechT-GL-EN— 48000 Hzjzsues/genshin-voice-zh— 48000 HzKishor798/text_to_speech_dataset— 48000 HzKppwdfgu1/Hypa-Speech-10k— 48000 HzKrorngAI/fleurs_openslr42_mpwt— 48000 HzKukedlc/openslr61-es-ar-full— 48000 Hzleduckhai/MultiMed— 48000 Hzleduckhai/MultiMed-ST— 48000 HzLeVy4/speech-to-text— 48000 Hzlilgoose777/nepali_speech_english_translation_shuffle_dataset— 48000 HzLindarychwalski/pony-speech— 48000 Hzlonghim99/khm-asr-cultural— 48000 Hzlyhourt-FSA/khm-asr-cultural— 48000 Hzmaikezu/dowis— 48000 Hzmanassehzw/sna-manasseh-150-raw— 48000 Hzmasuidrive/cv-corpus-17.0-zh-TW-client_id-grouped— 48000 HzMatrixStudio/TTS-CCabNavMSC— 48000 HzMatrixStudio/TTS-CFCabNavSC— 48000 HzMatrixStudio/TTS-SCCusSerFSC— 48000 HzMatrixStudio/TTS-SCDuFSC— 48000 HzMax5ive/openslr-32-hq-SA-languages-Sesotho— 48000 Hzmazesmazes/jenny-mimi— 48000 HzMikCil/f1-team-radio— 48000 Hzmohamedmou/DATASET-darija-ASR-clean— 48000 HzMohamedRashad/arabic-english-code-switching— 48000 HzMohamedRashad/common-voice-18-arabic— 48000 Hzngia/ASR_pulaar— 48000 Hzntaquan0125/steinsgate-voice— 48000 Hzntt123/VietBibleVox-aligned— 48000 HzOmarAhmedSobhy/egyption-with-emotion-dataset— 48000 HzOmarAhmedSobhy/tts-egyption-dataset— 48000 Hzomersaidd/tts_ahmet_deniz_tur— 48000 Hzrahafvii/EGY2K— 48000 Hzranbirchabungbam/meiteimayek-audio-parallel-corpus— 48000 Hzreapzor/neurologySTT— 48000 HzRikkaBotan/nyan-jenny-format— 48000 HzRobotsMali/transcription-scorer— 48000 HzSachinTelecmi/tts-hindi-stts2— 48000 Hzsartifyllc/Sukuma-Voices— 48000 Hzscriptaudio/f1-team-radio— 48000 Hzsdcsdccdsd/CSEMOTIONS— 48000 Hzshoron08/irodori-refs-10k— 48000 Hzshreeshacharya/Dhravani— 48000 Hzsiddiqiya/ar-eg-dataset— 48000 Hzslprl/StressTest— 48000 Hzsmcproject/MSC— 48000 Hzsomu9/iisc_mono_hindi_female— 48000 Hzsomu9/iitm_mono_hindi_female— 48000 Hzspeech-uk/opentts-kateryna— 48000 Hzspeech-uk/opentts-oleksa— 48000 Hzspeech-uk/opentts-tetiana— 48000 Hzspeech-uk/tts-crh-abibullah— 48000 Hzspeech-uk/tts-crh-arslan— 48000 Hzspeech-uk/tts-crh-sevil— 48000 HzSPRINGLab/IndicTTS-Hindi— 48000 HzSPRINGLab/IndicTTS_Assamese— 48000 HzSPRINGLab/IndicTTS_Bengali— 48000 HzSPRINGLab/IndicTTS_Kannada— 48000 HzSPRINGLab/IndicTTS_Malayalam— 48000 HzSPRINGLab/IndicTTS_Manipuri— 48000 HzSPRINGLab/IndicTTS_Marathi— 48000 HzSPRINGLab/IndicTTS_Odia— 48000 HzSPRINGLab/IndicTTS_Punjabi— 48000 HzSPRINGLab/IndicTTS_Rajasthani— 48000 HzSPRINGLab/IndicTTS_Tamil— 48000 HzSPRINGLab/IndicVoices-R_Hindi— 48000 Hzsrezas/farsi_voice_dataset— 48000 HzSynDataLab-JA-Refs/Irodori-Ja-Spk1-10k— 48000 HzSynDataLab-JA-Refs/Irodori-Ja-Spk2-10k— 48000 HzSynDataLab-JA-Refs/Irodori-Ja-Spk3-10k— 48000 HzSynDataLab-JA-Refs/Irodori-Ja-Spk4-10k— 48000 HzSynDataLab-JA-Refs/irodori-refs-10k— 48000 HzSynDataLab-JA-Refs/irodori-refs-10k-v2— 48000 HzSynDataLab-JA-Refs/irodori-tts-refs-12k— 48000 HzSynDataLab-JA/Irodori-Ja-500M-v2-vs-600M-v3-compare-20— 48000 Hzsynthbot/pony-singing— 48000 Hzsynthbot/pony-speech— 48000 HzTamazight-NLP/TOSD— 48000 Hzthennal/GMaSC— 48000 Hzthennal/indic_tts_ml— 48000 Hzthennal/msc— 48000 HzTrelis/multimed-hard— 48000 Hztrysem/indicvoices_r-ML— 48000 Hzttthe/MultiMed— 48000 Hztunis-ai/arabic_speech_corpus— 48000 Hzvinaybabu/voice_tech_for_all_challenge_samples_output— 48000 Hzvoice-biomarkers/openslr-32-hq-SA-languages-Afrikaans— 48000 Hzvoice-biomarkers/openslr-32-hq-SA-languages-isiXhosa— 48000 Hzvoice-biomarkers/openslr-32-hq-SA-languages-Sesotho— 48000 Hzvoice-biomarkers/openslr-32-hq-SA-languages-Setswana— 48000 Hzvrclc/openslr63— 48000 Hzwanasash/enwaucymraeg— 48000 Hzworldboss/ewe_bible_v2_tts— 48000 Hzworldboss/twi_bible_v2_tts— 48000 Hzyasalma/tat_hackathon_asr— 48000 HzYehor/qirimtatar-tts— 48000 Hzylacombe/english_dialects— 48000 Hzylacombe/google-chilean-spanish— 48000 Hzylacombe/google-tamil— 48000 Hzymoslem/CoVoST2-EN-AR— 48000 Hzymoslem/Living-Audio-Irish— 48000 Hzzinc75/Vibravox_dummy— 48000 Hz8Opt/clotho-dev-sample— 44100 Hzaangelakis/STOMA— 44100 Hzadiren7/darija_speech_to_text— 44100 Hzahmed220v/SCC22— 44100 HzAigizK/bashkort_tts_dataset— 44100 Hzaipanjab/speech-mendeley-pa— 44100 HzALEKAS/ToneBooksPlus-Grigorii— 44100 Hzalimetin/turkish-parliament-speech— 44100 Hzamine-khelif/DuBLaB-en-fr-0.7— 44100 Hzamine-khelif/DuBLaB-en-fr-0.7-f-0.2— 44100 Hzamine-khelif/DuBLaB-en-fr-0.7-f-0.5— 44100 Hzamine-khelif/DuBLaB-en-fr-0.8— 44100 HzAnilosan15/YouTube_Video_Transkriptleri_TR— 44100 HzAnonXx/Pidgin_ASR_Dataset_Combined— 44100 Hzanzorq/kbd_speech— 44100 Hzarchivartaunik/Jevanhielle_Zyhamont_outChecked— 44100 Hzarchivartaunik/output4Checked— 44100 HzBateesa/rw-tts-dataset— 44100 Hzberatcmn/jessica-076— 44100 Hzbezzam/coraal— 44100 HzBretagne/Lingua_Libre_br— 44100 HzBrunoHays/Bangor-Miami-Spanish-English-Corpus— 44100 Hzcagataydev/vlm-voice-audio— 44100 HzCentificAIResearch/DialectalSpeech-ICL— 44100 Hzchangelinglab/speechaccentarchive-pr— 44100 Hzchris-t-jansen/erasmian_greek_nt— 44100 HzCodyfederer/fttrtest— 44100 HzCodyfederer/test3434234— 44100 HzCodyfederer/tretret34543— 44100 HzCoRal-project/coral-tts— 44100 Hzcorti/med-term— 44100 Hzctaguchi/killkan— 44100 Hzczyzi0/pwr-azon-speech-dataset— 44100 Hzczyzi0/the-mc-speech-dataset— 44100 HzD00Movenok/russian-glados-portal2— 44100 Hzdaanbrugmans/ovb-huissen-1— 44100 Hzdata-lab-voice/echo-tts-en-benchmarks-v1— 44100 HzDataStudio/Vietnamese_ASR_TestingData_Old— 44100 HzEMINES/Tamazight-Speech-to-Arabic-Text— 44100 Hzfablevi/one_voice_FACEBOOK_PARQUET— 44100 Hzfarbodbij/persian-words— 44100 Hzfiifinketia/twi-trigrams-speech-text-parallel— 44100 Hzfosters/ales-krautsevich-karotkaia-gistoryia-belarusi-uladzimir-lisouski-output_original— 44100 Hzfosters/ales-krautsevich-karotkaia-gistoryia-belarusi-uladzimir-lisouski_all— 44100 Hzfosters/ales_razanau_all— 44100 Hzfosters/ales_razanau_output_original— 44100 Hzfosters/ales_zhuk_praklytaya_lyubow_all— 44100 Hzfosters/ales_zhuk_praklytaya_lyubow_output_original— 44100 Hzfosters/anatol_vyartsinski_pesnya_pra_hleb_all— 44100 Hzfosters/anatol_vyartsinski_pesnya_pra_hleb_output_original— 44100 Hzfosters/andre_marua_pakaranne_zolatam_all— 44100 Hzfosters/andre_marua_pakaranne_zolatam_output_original— 44100 Hzfosters/Antuan_de_Sent-Ekziupiery_Planieta_ludziei_all— 44100 Hzfosters/Antuan_de_Sent-Ekziupiery_Planieta_ludziei_output_original— 44100 Hzfosters/astryd_lindgren_braty_lvinae_sertsa_all— 44100 Hzfosters/astryd_lindgren_braty_lvinae_sertsa_output_original— 44100 Hzfosters/bely_klyck_all— 44100 Hzfosters/dzhozef_redzyard_kipling_all— 44100 Hzfosters/dzhozef_redzyard_kipling_output_original— 44100 Hzfosters/dzintra_shultse_robertsik_all— 44100 Hzfosters/dzintra_shultse_robertsik_output_original— 44100 Hzfosters/ernest_heminguei_stary_chalavek_i_mora_all— 44100 Hzfosters/ernest_heminguei_stary_chalavek_i_mora_output_original— 44100 Hzfosters/eryh_maryya_remark_all— 44100 Hzfosters/eryh_maryya_remark_output_original— 44100 Hzfosters/eryh_raspe_prygody_barona_myunhau_zena_all— 44100 Hzfosters/eryh_raspe_prygody_barona_myunhau_zena_output_original— 44100 Hzfosters/genadz_pashkou_all— 44100 Hzfosters/genadz_pashkou_output_original— 44100 Hzfosters/iagan_frydryh_shyler_kubak_all— 44100 Hzfosters/iagan_frydryh_shyler_kubak_output_original— 44100 Hzfosters/iakub-kolas-kazki-zhytstsia-output_original— 44100 Hzfosters/iakub-kolas-kazki-zhytstsia_all— 44100 Hzfosters/ivan-melezh-podykh-navalnitsy-valer-budzevich-output_original— 44100 Hzfosters/ivan-melezh-podykh-navalnitsy-valer-budzevich_all— 44100 Hzfosters/ivan-melezh-zavei-snezhan-valer-budzevich-output_original— 44100 Hzfosters/ivan-melezh-zavei-snezhan-valer-budzevich_all— 44100 Hzfosters/ivan-navumenka-zhul-vern-output_original— 44100 Hzfosters/ivan-navumenka-zhul-vern_all— 44100 Hzfosters/ivan-ptashnikau-lvy-output_original— 44100 Hzfosters/ivan-ptashnikau-lvy_all— 44100 Hzfosters/ivan_navumenka_sasna_pry_daroze_all— 44100 Hzfosters/ivan_navumenka_sasna_pry_daroze_output_original— 44100 Hzfosters/ivan_ptashnikau_all— 44100 Hzfosters/ivan_ptashnikau_output_original— 44100 Hzfosters/ivan_shamyakin_sertsa_na_daloni_all— 44100 Hzfosters/ivan_shamyakin_sertsa_na_daloni_output_original— 44100 Hzfosters/ivan_shamyakin_tryvozhnae_shchastse_all— 44100 Hzfosters/ivan_shamyakin_tryvozhnae_shchastse_output_original— 44100 Hzfosters/knihi-be-arlou_tancy_nad_horadam_all— 44100 Hzfosters/knihi-be-arlou_tancy_nad_horadam_output_original— 44100 Hzfosters/kuzma_chorny_makarkavyh_volka_all— 44100 Hzfosters/kuzma_chorny_makarkavyh_volka_output_original— 44100 Hzfosters/kuzma_chorny_poshuki_buduchyni_all— 44100 Hzfosters/kuzma_chorny_poshuki_buduchyni_output_original— 44100 Hzfosters/kuzma_chorny_zyamlya_all— 44100 Hzfosters/kuzma_chorny_zyamlya_output_original— 44100 Hzfosters/legendy-i-padanni_all— 44100 Hzfosters/legendy-i-padanni_original— 44100 Hzfosters/maksim_tank_all— 44100 Hzfosters/maksim_tank_output_original— 44100 Hzfosters/mar_yan_duksa_all— 44100 Hzfosters/mar_yan_duksa_output_original— 44100 Hzfosters/raisa_baravikova_vasmiradkou_i_all— 44100 Hzfosters/raisa_baravikova_vasmiradkou_i_output_original— 44100 Hzfosters/raisa_baravikova_vershy_pra_kahanne_all— 44100 Hzfosters/raisa_baravikova_vershy_pra_kahanne_output_original— 44100 Hzfosters/shata_rustaveli_vitsyaz_u_tygravai_shkury_all— 44100 Hzfosters/shata_rustaveli_vitsyaz_u_tygravai_shkury_output_original— 44100 Hzfosters/stefan_tsvei_g_nyabachnaya_kalektsyya_all— 44100 Hzfosters/stefan_tsvei_g_nyabachnaya_kalektsyya_output_original— 44100 Hzfosters/taras_shau_chenka_vershy_paemy_all— 44100 Hzfosters/taras_shau_chenka_vershy_paemy_output_original— 44100 Hzfosters/uilyam_folkner_pah_verbeny_all— 44100 Hzfosters/uilyam_folkner_pah_verbeny_output_original— 44100 Hzfosters/uladzimir-karatkevich-dzikae-paliavanne-karalia-stakha-aleg-garbuz-output_original— 44100 Hzfosters/uladzimir-karatkevich-dzikae-paliavanne-karalia-stakha-aleg-garbuz_all— 44100 Hzfosters/vasil_bykau_all— 44100 Hzfosters/vasil_bykau_output_original— 44100 Hzfosters/vasil_zue_nok_syaliba_all— 44100 Hzfosters/vasil_zue_nok_syaliba_output_original— 44100 Hzfosters/viktar_prau_dzin_all— 44100 Hzfosters/viktar_prau_dzin_output_original— 44100 Hzfosters/yakub_kolas_novaya_zyamlya_all— 44100 Hzfosters/yakub_kolas_novaya_zyamlya_output_original— 44100 Hzfosters/yanka_bryl_ptushki_i_gne_zdy_all— 44100 Hzfosters/yanka_bryl_ptushki_i_gne_zdy_output_original— 44100 Hzfosters/yanka_sipakou_odzium_all— 44100 Hzfosters/yanka_sipakou_odzium_output_original— 44100 HzfutureDoctor/turkic_tts_dataset— 44100 Hzghananlpcommunity/asante-twi-bible-speech-phonemes— 44100 Hzghananlpcommunity/twi-trigrams-speech-text-parallel— 44100 Hzgrandhigh/sample-id— 44100 Hzgrider-transwithai/nekopara-speech— 44100 Hzhananeek2/STT-algerian-dialect— 44100 Hzhhim8826/japanese-anime-speech-v2-split— 44100 Hzhumairawan/AnimeSpeech— 44100 Hzjdapaah/asante-twi-bible— 44100 Hzjoujiboi/kuroyukihime-speech— 44100 Hzkizuna-intelligence/AItuber-Persona-Voices-JA— 44100 Hzlangswap/dialogs-ru-emotional-conversations— 44100 HzMatrixStudio/TTS-SCFChilSC— 44100 Hzmetricv/tl-whisper— 44100 HzMichel21/rick-sanchez— 44100 HzMikhailT/hifi-tts— 44100 HzMohamedRashad/SCC22— 44100 HzMothersTongue/mother_tongue_dataset— 44100 Hzmsnowchanj/pvariant-EQ— 44100 HzPragmaticl/Trys2— 44100 HzPragmaticl/TuyenVanHoa2— 44100 HzPragmaticl/TuyenVanHoa4— 44100 HzProfessor/kinyarwanda-tts-dataset-kin— 44100 Hzray0rf1re/GLaDOS-audio-v2— 44100 Hzsachin6624/malayalam-tts-pro-voice— 44100 Hzshannonnonshan/ViMedCSS-Cop— 44100 HzShoukanLabs/AniSpeech— 44100 HzSoufianeDahimi/Tamazight-ASR-Dataset-v2— 44100 Hzsudoping01/bam-asr-benchmark— 44100 HzSynDataLab-EN-Refs/echo-ref-speakers-4k-en— 44100 HzSynDataLab-EN-Refs/tts-pretrain-refs-3k-mos— 44100 HzSynDataLab-EN/EchoTTS-OmniVoice-En— 44100 Hztensorxt/ViMedCSS— 44100 HzThorsten-Voice/TV-44kHz-Full— 44100 Hztimniel/Pidgin_ASR_Dataset_Combined— 44100 Hztrysem/malayalam-tts-pro-voice— 44100 HzTutlaytAI/kabyle_asr— 44100 HzTutlaytAI/Kabyle_ASR-En_Translation— 44100 HzTutlaytAI/Kabyle_ASR-Fr_Translation— 44100 Hzvsisik/voice-dataset-lili— 44100 Hzwcwxyz/test-audio— 44100 HzWhissleAI/Meta_STT_ZH_AIShell3— 44100 Hzyasalma/audiobooks— 44100 Hzymoslem/BitesizeIrish-GA-EN— 44100 Hzyuriilaba/toronto-tv-ukrainian— 44100 Hz