jedisct1/Arch-Router-1.5B

1---2base_model: katanemo/Arch-Router-1.5B3language:4- en5library_name: mlx6license: other7license_name: katanemo-research8license_link: https://huggingface.co/katanemo/Arch-Router-1.5B/blob/main/LICENSE9pipeline_tag: text-generation10tags:11- routing12- preference13- arxiv:2506.1665514- llm15- mlx16paper: https://arxiv.org/abs/2506.1665517---18 19# jedisct1/Arch-Router-1.5B20 21This model [jedisct1/Arch-Router-1.5B](https://huggingface.co/jedisct1/Arch-Router-1.5B) was22converted to MLX format from [katanemo/Arch-Router-1.5B](https://huggingface.co/katanemo/Arch-Router-1.5B)23using mlx-lm version **0.25.2**.24 25## Use with mlx26 27```bash28pip install mlx-lm29```30 31```python32from mlx_lm import load, generate33 34model, tokenizer = load("jedisct1/Arch-Router-1.5B")35 36prompt = "hello"37 38if tokenizer.chat_template is not None:39    messages = [{"role": "user", "content": prompt}]40    prompt = tokenizer.apply_chat_template(41        messages, add_generation_prompt=True42    )43 44response = generate(model, tokenizer, prompt=prompt, verbose=True)45```46