library_name: transformers base_model: deepseek-ai/DeepSeek-R1-Distill-Qwen-32B tags:

• mlx

mlx-community/DeepSeek-R1-Distill-Qwen-32B-4bit

模型 mlx-community/DeepSeek-R1-Distill-Qwen-32B-4bit 是从 deepseek-ai/DeepSeek-R1-Distill-Qwen-32B 转换而来的 MLX 格式，使用了 0.20.2 版本的 mlx-lm。

在 MLX 中使用

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("mlx-community/DeepSeek-R1-Distill-Qwen-32B-4bit")

prompt="你好"

if hasattr(tokenizer, "apply_chat_template") and tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, tokenize=False, add_generation_prompt=True
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)