Instructions to use tinywell/THUDM-glm-4-9b-chat-4bit with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use tinywell/THUDM-glm-4-9b-chat-4bit with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir THUDM-glm-4-9b-chat-4bit tinywell/THUDM-glm-4-9b-chat-4bit
- Notebooks
- Google Colab
- Kaggle
- Local Apps
- LM Studio
how did you make mlx_lm support glm-4?
#2
by oldhu - opened
When I try to convert using mlx_lm.convert, it says chatglm is not supported.
When I try to convert using mlx_lm.convert, it says chatglm is not supported.
I rewrote the chatglm model using mlx, so I can perform parameter conversion (chatglm.py). However, there are some problems with the inferencing, which are still being debugged.