MakeDiffSinger
Pipelines and tools to build your own DiffSinger dataset.
Install / Use
/learn @openvpi/MakeDiffSingerREADME
MakeDiffSinger
Pipelines and tools to build your own DiffSinger dataset.
For the recommended standard dataset making pipelines, see:
- acoustic-forced-alignment: make dataset from scratch with MFA for acoustic model training
- variance-temp-solution: temporary solution to extend acoustic datasets into variance datasets
For other useful pipelines and tools for making a dataset, welcome to raise issues or submit PRs.
DiffSinger dataset structure
- dataset1/
- raw/
- wavs/
- recording1.wav
- recording2.wav
- ...
- transcriptions.csv
- wavs/
- raw/
- dataset2/
- raw/
- wavs/
- ...
- transcriptions.csv
- wavs/
- raw/
- ...
Essential tools to process and label your datasets
Dataset tools now have their own repository: dataset-tools.
There are mainly 3 components:
- AudioSlicer: Slice your recordings into short segments
- MinLabel: Label *.lab files containing word transcriptions for acoustic model training.
- SlurCutter: Edit MIDI sequence in *.ds files for variance model training.
Related Skills
node-connect
345.9kDiagnose OpenClaw node connection and pairing failures for Android, iOS, and macOS companion apps
frontend-design
106.4kCreate distinctive, production-grade frontend interfaces with high design quality. Use this skill when the user asks to build web components, pages, or applications. Generates creative, polished code that avoids generic AI aesthetics.
openai-whisper-api
345.9kTranscribe audio via OpenAI Audio Transcriptions API (Whisper).
qqbot-media
345.9kQQBot 富媒体收发能力。使用 <qqmedia> 标签,系统根据文件扩展名自动识别类型(图片/语音/视频/文件)。
