README.md

metadata

base_model: sm54/QwQ-Qwen2.5-Coder-Instruct-32B
library_name: transformers
tags:
  - mergekit
  - merge
  - mlx
  - mlx-my-repo

bobig/QwQ-Qwen2.5-Coder-Instruct-32B-4bit

The Model bobig/QwQ-Qwen2.5-Coder-Instruct-32B-4bit was converted to MLX format from sm54/QwQ-Qwen2.5-Coder-Instruct-32B using mlx-lm version 0.21.5.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("bobig/QwQ-Qwen2.5-Coder-Instruct-32B-4bit")

prompt="hello"

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)