cnfusion/MusaCoder-27B-mlx-3Bit

1---2license: apache-2.03language:4- en5- zh6base_model: MooreThreads/MusaCoder-27B7pipeline_tag: reinforcement-learning8tags:9- CUDA10- MUSA11- GPU-Kernel12- Reinforcement-Learning13- mlx14- mlx-my-repo15---16 17# cnfusion/MusaCoder-27B-mlx-3Bit18 19The Model [cnfusion/MusaCoder-27B-mlx-3Bit](https://huggingface.co/cnfusion/MusaCoder-27B-mlx-3Bit) was converted to MLX format from [MooreThreads/MusaCoder-27B](https://huggingface.co/MooreThreads/MusaCoder-27B) using mlx-lm version **0.31.2**.20 21## Use with mlx22 23```bash24pip install mlx-lm25```26 27```python28from mlx_lm import load, generate29 30model, tokenizer = load("cnfusion/MusaCoder-27B-mlx-3Bit")31 32prompt="hello"33 34if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:35    messages = [{"role": "user", "content": prompt}]36    prompt = tokenizer.apply_chat_template(37        messages, tokenize=False, add_generation_prompt=True38    )39 40response = generate(model, tokenizer, prompt=prompt, verbose=True)41```42