README.md

metadata

license: other
license_name: yandexgpt-5-lite-8b-pretrain
license_link: LICENSE
language:
  - ru
  - en
base_model: yandex/YandexGPT-5-Lite-8B-pretrain
tags:
  - mlx

mlx-community/YandexGPT-5-Lite-8B-pretrain-Q8-mlx

The Model mlx-community/YandexGPT-5-Lite-8B-pretrain-Q8-mlx was converted to MLX format from yandex/YandexGPT-5-Lite-8B-pretrain using mlx-lm version 0.20.5.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("mlx-community/YandexGPT-5-Lite-8B-pretrain-Q8-mlx")

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)