Starstruk Logo MarkStarstruk

source-separation-models

k2-fsa/sherpa-onnxsource-separation-modelsMay 22, 2025by csukuangfj

AI Summary

This release provides pre-converted audio samples and model weights for Spleeter and UVR source separation models.

Key Highlights

  • UVR models converted from TRvlvr repo
  • Spleeter audio example

New Features

  • Pre-trained UVR models
  • Pre-trained Spleeter models

Full Release Notes

`audio_example.wav` is converted from
https://github.com/deezer/spleeter/blob/v1.4.0/audio_example.mp3
with the following command
```bash
sox audio_example.mp3 audio_example.wav
```

UVR models are converted from
https://github.com/TRvlvr/model_repo/releases/tag/all_public_uvr_models
using the code from
https://github.com/k2-fsa/sherpa-onnx/tree/master/scripts/uvr_mdx