README.md

metadata

license: mit

b150_official_train

This release contains the high-confidence b150 training split used by MAGIC-TTS, with standalone MFA word-level alignments distributed separately from prepared Arrow artifacts.

Related links:

• GitHub repo: https://github.com/yongaifadian1/MAGIC-TTS
• arXiv paper: https://arxiv.org/abs/2604.21164
• Open-source model: https://huggingface.co/maimai11/MAGIC-TTS
• Online demo: https://yongaifadian1.github.io/MAGIC-TTS/

Stats:

• Total samples: 201986
• English: 78363
• Chinese: 123623

Files:

• selected_samples.exported.jsonl: manifest with sample_id, utt_id, language, audio_path, duration, and target_text
• raw_audio_en.tar.part-*: split English raw audio bundle
• raw_audio_zh.tar.part-*: split Chinese raw audio bundle
• mfa_sidecars.tar.part-*: split standalone MFA word-level alignments with relative audio_path, ready to be reused during local dataset preparation

Notes:

• Reconstruct tar files first:

cat raw_audio_en.tar.part-* > raw_audio_en.tar
cat raw_audio_zh.tar.part-* > raw_audio_zh.tar
cat mfa_sidecars.tar.part-* > mfa_sidecars.tar

• Then extract:

mkdir -p raw
tar -xf raw_audio_en.tar -C raw
tar -xf raw_audio_zh.tar -C raw
tar -xf mfa_sidecars.tar

• After extraction, the dataset layout becomes raw/audio/... and mfa_sidecars/....
• Arrow artifacts are intentionally not included in this release.