---
title: Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx
canonical_url: "https://www.modelscope.cn/models/mlx-community/Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx"
md_url: "https://www.modelscope.cn/models/mlx-community/Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx.md"
repository: mlx-community/Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx
last_updated: 2026-05-05
license: other
model_type:
  - nemotron
architectures:
  - NemotronForCausalLM
base_model:
  - nvidia/Nemotron-Mini-4B-Instruct
base_model_relation: finetune
parameters: 4.2B
tensor_type:
  - BF16
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
language:
  - en
downloads: 5
stars: 0
tags:
  - mlx
  - llm
  - nemotron
  - apple-silicon
---

# Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx

> Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx - mlx-community 在 ModelScope 开源的模型。Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx

mlx-community/Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx 是 ModelScope 魔搭社区上的 4.2B 参数机器学习模型，采用 other 许可，基于 nvidia/Nemotron-Mini-4B-Instruct 构建。

- **Repository**: mlx-community/Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx
- **License**: other
- **Parameters**: 4.2B
- **Base model**: nvidia/Nemotron-Mini-4B-Instruct
- **Tags**: mlx, llm, nemotron, apple-silicon
- **Downloads**: 5
- **Stars**: 0
- **Last updated**: 2026-05-05

Source: https://www.modelscope.cn/models/mlx-community/Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx

---

# Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx

This model was converted from [nvidia/Nemotron-Mini-4B-Instruct](https://huggingface.co/nvidia/Nemotron-Mini-4B-Instruct)
to [MLX](https://github.com/ml-explore/mlx) format for use on Apple Silicon.

**Quantization:** 4-bit mxfp4 quantization

## Usage

```python
from mlx_lm import load, generate

model, tokenizer = load("mlx-community/Nemotron-Mini-4B-Instruct-mxfp4-4bit-mlx")

prompt = (
    "<extra_id_0>System\n"
    "You are a helpful, honest AI assistant.\n\n"
    "<extra_id_1>User\n"
    "Who are you?\n"
    "<extra_id_1>Assistant\n"
)

print(generate(model, tokenizer, prompt, max_tokens=256))
```

## Benchmark (Apple Silicon, single prompt, 23 tokens)

| Variant | tok/s |
|---|---|
| bf16 (this) | 2.47 |
| 4-bit default | 4.37 |
| mxfp4-q4 | 4.56 |
| nvfp4-q4 | 9.69 |
| mixed-3-6 | 9.72 |

## Original model

See [nvidia/Nemotron-Mini-4B-Instruct](https://huggingface.co/nvidia/Nemotron-Mini-4B-Instruct)
for the original model card, license, and usage terms.