KAT-Dev-qx86-hi-mlx

Great Coding model, a bit slow. The qx65-hi quant is faster.

Running tests on it will take a while, initial metrics show high performance

This model KAT-Dev-qx86-hi-mlx was converted to MLX format from Kwaipilot/KAT-Dev using mlx-lm version 0.28.0.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("KAT-Dev-qx86-hi-mlx")

prompt = "hello"

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

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