Instructions to use Jundot/GLM-5.2-oQ4e-mtp with libraries, inference providers, notebooks, and local apps. Follow these links to get started.
- Libraries
- MLX
How to use Jundot/GLM-5.2-oQ4e-mtp with MLX:
# Download the model from the Hub pip install huggingface_hub[hf_xet] huggingface-cli download --local-dir GLM-5.2-oQ4e-mtp Jundot/GLM-5.2-oQ4e-mtp
- Notebooks
- Google Colab
- Kaggle
- Local Apps Settings
- LM Studio
- Atomic Chat
Benchmarks on M3 Ultra 512GB (oMLX v0.5.1): MTP +33% decode, quality vs imatrix GGUF
👍 1
1
#1 opened about 1 month ago
by
WillMitchell