---
title: Mistral-7B-Instruct-v0.3-mlx-4Bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/Mistral-7B-Instruct-v0.3-mlx-4Bit"
md_url: "https://www.modelscope.cn/models/mlx-community/Mistral-7B-Instruct-v0.3-mlx-4Bit.md"
repository: mlx-community/Mistral-7B-Instruct-v0.3-mlx-4Bit
last_updated: 2026-07-27
license: apache-2.0
model_type:
  - mistral
architectures:
  - MistralForCausalLM
base_model:
  - mistralai/Mistral-7B-Instruct-v0.3
base_model_relation: quantized
parameters: 1.1B
tensor_type:
  - BF16
  - U32
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 11
stars: 0
tags:
  - vllm
  - mistral-common
  - mlx
  - mlx-my-repo
---

# Mistral-7B-Instruct-v0.3-mlx-4Bit

> Mistral-7B-Instruct-v0.3-mlx-4Bit - mlx-community 在 ModelScope 开源的模型。SkyStar-tech/Mistral-7B-Instruct-v0.3-mlx-4Bit

mlx-community/Mistral-7B-Instruct-v0.3-mlx-4Bit 是 ModelScope 魔搭社区上的 1.1B 参数机器学习模型，采用 apache-2.0 许可，基于 mistralai/Mistral-7B-Instruct-v0.3 构建。

- **Repository**: mlx-community/Mistral-7B-Instruct-v0.3-mlx-4Bit
- **License**: apache-2.0
- **Parameters**: 1.1B
- **Base model**: mistralai/Mistral-7B-Instruct-v0.3
- **Tags**: vllm, mistral-common, mlx, mlx-my-repo
- **Downloads**: 11
- **Stars**: 0
- **Last updated**: 2026-07-27

Source: https://www.modelscope.cn/models/mlx-community/Mistral-7B-Instruct-v0.3-mlx-4Bit

---

# SkyStar-tech/Mistral-7B-Instruct-v0.3-mlx-4Bit

The Model [SkyStar-tech/Mistral-7B-Instruct-v0.3-mlx-4Bit](https://huggingface.co/SkyStar-tech/Mistral-7B-Instruct-v0.3-mlx-4Bit) was converted to MLX format from [mistralai/Mistral-7B-Instruct-v0.3](https://huggingface.co/mistralai/Mistral-7B-Instruct-v0.3) using mlx-lm version **0.31.2**.

## Use with mlx

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

```python
from mlx_lm import load, generate

model, tokenizer = load("SkyStar-tech/Mistral-7B-Instruct-v0.3-mlx-4Bit")

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)
```