Instructions to use mlx-community/gemma-4-e4b-it-mxfp4 with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/gemma-4-e4b-it-mxfp4 with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir gemma-4-e4b-it-mxfp4 mlx-community/gemma-4-e4b-it-mxfp4
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
Leaks metadata into chat
#3
by rajo82 - opened
I'm not sure where the issue comes from but as no other model does it, I assume it's the model. Using oMLX 0.3.10 (and 0.3.9), most chats with the model result in garbage leaking into the chat like this.
Tried, didn't fix:
- deleting and reinstalling the mode
- restarting computer and/or the AI stack (Open Webui 0.9.5 in Docker, oMLX 0.3.9)
- deleting the Workspace Model, calling the underlying model directly
