from mlx_lm import load, generate
from huggingface_hub import snapshot_download

path = snapshot_download(
    repo_id="arcee-ai/Arcee-Maestro-7B-Preview-MLX",
    allow_patterns="4bit/*",  # This will download everything in the 4bit folder
    local_dir="Arcee-Maestro-7B-Preview-4bit"    # Optional: specify where to save
)

model, tokenizer = load("Arcee-Maestro-7B-Preview-4bit")

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)