README.md

metadata

language:
  - ja
  - en
license: apache-2.0
tags:
  - japanese
  - causal-lm
  - mlx
inference: false

mlx-community/calm3-22b-chat-4bit

The Model mlx-community/calm3-22b-chat-4bit was converted to MLX format from cyberagent/calm3-22b-chat using mlx-lm version 0.15.2.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("mlx-community/calm3-22b-chat-4bit")
response = generate(model, tokenizer, prompt="hello", verbose=True)