README.md

metadata

license: apache-2.0
language:
  - en
  - ko
base_model: sosoai/hansoldeco-qwen2.5-grpo-gsm-8k-ko-v1-mlx
tags:
  - mlx
  - mlx
  - mlx-my-repo

sosoai/hansoldeco-qwen2.5-grpo-gsm-8k-ko-v1-mlx-Q2-mlx

The Model sosoai/hansoldeco-qwen2.5-grpo-gsm-8k-ko-v1-mlx-Q2-mlx was converted to MLX format from sosoai/hansoldeco-qwen2.5-grpo-gsm-8k-ko-v1-mlx using mlx-lm version 0.21.5.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("sosoai/hansoldeco-qwen2.5-grpo-gsm-8k-ko-v1-mlx-Q2-mlx")

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)