---
title: Seed-OSS-36B-Instruct-8bit
canonical_url: "https://www.modelscope.cn/models/mlx-community/Seed-OSS-36B-Instruct-8bit"
md_url: "https://www.modelscope.cn/models/mlx-community/Seed-OSS-36B-Instruct-8bit.md"
repository: mlx-community/Seed-OSS-36B-Instruct-8bit
last_updated: 2025-09-03
license: apache-2.0
pipeline_tag: text-generation
tasks:
  - text-generation
model_type:
  - seed_oss
architectures:
  - SeedOssForCausalLM
base_model:
  - ByteDance-Seed/Seed-OSS-36B-Instruct
base_model_relation: quantized
parameters: 10.2B
tensor_type:
  - U32
  - BF16
library_name:
  - mlx
  - safetensors
  - pytorch
frameworks:
  - pytorch
downloads: 107
stars: 1
tags:
  - vllm
  - mlx
  - mlx-my-repo
---

# Seed-OSS-36B-Instruct-8bit

> Seed-OSS-36B-Instruct-8bit - mlx-community 在 ModelScope 开源的模型。mlx-community/Seed-OSS-36B-Instruct-mlx-8Bit

mlx-community/Seed-OSS-36B-Instruct-8bit 是 ModelScope 魔搭社区上的 10.2B 参数text-generation模型，采用 apache-2.0 许可，基于 ByteDance-Seed/Seed-OSS-36B-Instruct 构建。

- **Repository**: mlx-community/Seed-OSS-36B-Instruct-8bit
- **License**: apache-2.0
- **Tasks**: text-generation
- **Parameters**: 10.2B
- **Base model**: ByteDance-Seed/Seed-OSS-36B-Instruct
- **Tags**: vllm, mlx, mlx-my-repo
- **Downloads**: 107
- **Stars**: 1
- **Last updated**: 2025-09-03

Source: https://www.modelscope.cn/models/mlx-community/Seed-OSS-36B-Instruct-8bit

---

# mlx-community/Seed-OSS-36B-Instruct-mlx-8Bit

The Model [mlx-community/Seed-OSS-36B-Instruct-mlx-8Bit](https://huggingface.co/mlx-community/Seed-OSS-36B-Instruct-mlx-8Bit) was converted to MLX format from [ByteDance-Seed/Seed-OSS-36B-Instruct](https://huggingface.co/ByteDance-Seed/Seed-OSS-36B-Instruct) using mlx-lm version **0.26.4**.

## Use with mlx

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

```python
from mlx_lm import load, generate

model, tokenizer = load("mlx-community/Seed-OSS-36B-Instruct-mlx-8Bit")

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)
```