Audio-to-Audio
MLX
audio
music-source-separation
source-separation
demucs
htdemucs
hdemucs
apple-silicon
Instructions to use iky1e/demucs-mlx with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use iky1e/demucs-mlx with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download iky1e/demucs-mlx --local-dir demucs-mlx
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Commit History
-
Replace demucs-mlx conversion scripts with direct PyTorch exporter
Kyle HowellsRemoves dependency on demucs-mlx re-implementation. The new export_from_pytorch.py converts all 8 models directly from the original PyTorch demucs package to safetensors + JSON config.3ff2e58 -
Update README: add model descriptions, update conversion path to direct PyTorch export
Kyle Howellsc4137e6 -
Add all 8 pretrained Demucs models and export scripts
Kyle Howells +1: Claude Opus 4.6Add safetensors weights and config JSON for all pretrained models: htdemucs, htdemucs_ft, htdemucs_6s, hdemucs_mmi, mdx, mdx_extra, mdx_q, mdx_extra_q. Include batch export scripts (export_all_models.py, export_mdx.py) and update README with full model table and usage instructions. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>0ef541a
-
Add htdemucs MLX weights and model card
Kyle Howellsa150c61 -
initial commit
Verified signaturef1d722d