README.md

---
base_model: deepseek-ai/DeepSeek-V2.5
license: other
license_name: deepseek
license_link: https://github.com/deepseek-ai/DeepSeek-V2/blob/main/LICENSE-MODEL
tags:
- mlx
---

# mlx-community/DeepSeek-V2.5-MLX-AQ4_1_64

The Model [mlx-community/DeepSeek-V2.5-MLX-AQ4_1_64](https://huggingface.co/mlx-community/DeepSeek-V2.5-MLX-AQ4_1_64) was converted to MLX format from [deepseek-ai/DeepSeek-V2.5](https://huggingface.co/deepseek-ai/DeepSeek-V2.5) using mlx-lm version **0.18.2**.

## Use with mlx

```bash
pip install mlx-lm
```

```python
from mlx_lm import load, generate

model, tokenizer = load("mlx-community/DeepSeek-V2.5-MLX-AQ4_1_64")

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)
```