typealias/Llama-3-6B-Instruct-pruned-mlx-4bit

The Model typealias/Llama-3-6B-Instruct-pruned-mlx-4bit was converted to MLX format from kuotient/Llama-3-6B-Instruct-pruned using mlx-lm version 0.13.0.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("typealias/Llama-3-6B-Instruct-pruned-mlx-4bit")
response = generate(model, tokenizer, prompt="hello", verbose=True)