---
library_name: mlx
model_name: Llama 3.3 Shisa V2.1 70B
license: llama3.3
pipeline_tag: text-generation
language:
- ja
- en
tags:
- transformers
- mlx
- translation
base_model:
- shisa-ai/shisa-v2.1-llama3.3-70b
datasets:
- shisa-ai/shisa-v2.1-sharegpt
---

# mlx-community/shisa-v2.1-llama3.3-70b-mlx-bf16

The Model [mlx-community/shisa-v2.1-llama3.3-70b-mlx-bf16](https://huggingface.co/mlx-community/shisa-v2.1-llama3.3-70b-mlx-bf16) was converted to MLX format from [shisa-ai/shisa-v2.1-llama3.3-70b](https://huggingface.co/shisa-ai/shisa-v2.1-llama3.3-70b) using mlx-lm version **0.28.4**.

You can find other similar translation-related MLX model quants for an Apple Mac at https://huggingface.co/bibproj

## Use with mlx

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

```python
from mlx_lm import load, generate

model, tokenizer = load("mlx-community/shisa-v2.1-llama3.3-70b-mlx-bf16")

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)
```