Chinese R18 ASMR audio dataset for TTS/voice cloning fine-tuning.
Works: 5876 RJ-coded works
Total size: ~1.1TB raw (10 compressed parts)
Format: MP3/WAV/FLAC, 3 tracks per work
Source: asmr.one API, Chinese R18 category
for f in packs/*.tar.zst; do
zstd -d "$f" --stdout | tar -xf -
done
