bobachicken/Llama-3.2-1B-Alpaca

The Model bobachicken/Llama-3.2-1B-Alpaca was finetuned using the Dataset bobachicken/alpaca-split , a split version of the Alpaca Dataset tatsu-lab/alpaca using the MLX Framework(https://github.com/ml-explore) developed by Apple. The fine-tuning process was conducted using DORA(https://developer.nvidia.com/blog/introducing-dora-a-high-performing-alternative-to-lora-for-fine-tuning/) for a total of 2000 iterations.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("bobachicken/Llama-3.2-1B-Alpaca")

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)