README.md

metadata

library_name: transformers
model_name: Vikhrmodels/Vikhr-YandexGPT-5-Lite-8B-it
base_model: Vikhrmodels/Vikhr-YandexGPT-5-Lite-8B-it
language:
  - ru
  - en
license: other
license_name: yandexgpt-5-lite-8b-pretrain
license_link: LICENSE
tags:
  - mlx

Vikhrmodels/Vikhr-YandexGPT-5-Lite-8B-it_MLX-8bit

The Model Vikhrmodels/Vikhr-YandexGPT-5-Lite-8B-it_MLX-8bit was converted to MLX format from Vikhrmodels/Vikhr-YandexGPT-5-Lite-8B-it using mlx-lm version 0.21.5.

Use with mlx

pip install mlx-lm

from mlx_lm import load, generate

model, tokenizer = load("Vikhrmodels/Vikhr-YandexGPT-5-Lite-8B-it_MLX-8bit")

prompt = "hello"

if tokenizer.chat_template is not None:
    messages = [{"role": "user", "content": prompt}]
    prompt = tokenizer.apply_chat_template(
        messages, add_generation_prompt=True
    )

response = generate(model, tokenizer, prompt=prompt, verbose=True)