Instructions to use mlx-community/Nous-Hermes-2-Mixtral-8x7B-DPO-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use mlx-community/Nous-Hermes-2-Mixtral-8x7B-DPO-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir Nous-Hermes-2-Mixtral-8x7B-DPO-4bit mlx-community/Nous-Hermes-2-Mixtral-8x7B-DPO-4bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Changement by mlx_lm 0.0.10 update lead to an error using the load, generate call
Hello,
I recently ran into a typeError (raise TypeError(f'non-default argument {f.name!r} ' TypeError: non-default argument 'model_type' follows default argument
) after updating the package.
Since im pretty new to all this im not really sure if this error is caused by myself, but apparently it seems like the changed file "mixtral.py" in mlx_lm/models and its dataclass-decorated class "ModelArgs(BaseModelArgs):" ist causing the error by the misplaced attribute "model_type: str". After I positioned it before the optional attributes, it worked again.
@BenBenser Good catch. Created PR here in the repository: https://github.com/ml-explore/mlx-examples/pull/450.
Thanks! And thanks for creating PR repo, wasn't sure if I should create one.