Modelpublic

Niwahi/Flux-Japanese-Qwen2.5-32B-Instruct-V1.0-mlx-4Bit

sourceHugging Faceapache-2.0updated 10mo agoView on Hugging Face

0likes175downloads

Model Card

Niwahi/Flux-Japanese-Qwen2.5-32B-Instruct-V1.0-mlx-4Bit

The Model Niwahi/Flux-Japanese-Qwen2.5-32B-Instruct-V1.0-mlx-4Bit was converted to MLX format from flux-inc/Flux-Japanese-Qwen2.5-32B-Instruct-V1.0 using mlx-lm version 0.28.3.

Use with mlx

bash

pip install mlx-lm

python

from mlx_lm import load, generate

model, tokenizer = load("Niwahi/Flux-Japanese-Qwen2.5-32B-Instruct-V1.0-mlx-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)