Hugging Face's logo

Datasets:
Tohirju
/
kazakh-tts-diarized

Tasks:
Text-to-Speech
Languages:
Kazakh
Tags:
kk
tts
diarization
voice-cloning
reference-bank
License:
Dataset card Data Studio Files
xet
 Community

You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

kk TTS diarization + voice-cloning reference bank (PUBLIC + gated=manual)

Diarized speaker labels + per-speaker enrollment reference bank for kk TTS / voice cloning, derived from the kk ASR corpora. ECAPA-clustered (16 kHz).

  • manifest.jsonl : 402176 chunks {audio_filepath(local), text, snr, duration, sr, spk_id, source}
  • ref_bank.jsonl : 3142 enrollment clips (3-10s, cleanest per speaker) — the voice-cloning bank
  • speaker_stats.csv : 1057 speakers (n_clips, total_min, mean_snr)
  • ecapa.npy : ECAPA speaker embeddings NOTE: audio wavs stay on the training box (derived from the ASR corpora already on HF). Feeds the CosyVoice2 flagship FT + cloning bank.
Downloads last month
3
Total file size:
447 MB