Instructions to use mgcrea/pupitre-models with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mgcrea/pupitre-models with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] hf download mgcrea/pupitre-models --local-dir pupitre-models
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Pupitre models
Models converted for Pupitre, a Mac app for practising piano with songs you like. Each folder holds one model with its own README: what it is, where the weights come from, their licence and the credit they need, the exact conversion command, and how the conversion was checked against the original.
| Folder | Model | Format | Precision | Size |
|---|---|---|---|---|
htdemucs-6s-mlx |
Meta's Hybrid Transformer Demucs, 6 stems (drums, bass, other, vocals, guitar, piano) | MLX safetensors | FP32 | 110 MB |
htdemucs-6s-fp32 |
The same model as a Core ML package, no longer used by Pupitre | Core ML | FP32 | 355 MB |
The app downloads these files pinned to a commit and checks each one's size and SHA-256.
- Downloads last month
- 10
Hardware compatibility
Log In to add your hardware
Quantized
Inference Providers NEW
This model isn't deployed by any Inference Provider. ๐ Ask for provider support