---
license: mit
base_model:
- deepseek-ai/DeepSeek-R1-Zero
---

# DeepSeek-R1-Zero-256x21B-BF16  

## Files Available  

**Imatrix:** [DeepSeek-R1-Zero-256x21B-BF16.imatrix](https://huggingface.co/gghfez/DeepSeek-R1-Zero-256x21B-BF16/blob/main/DeepSeek-R1-Zero.imatrix)  

**GGUF files:** Moved to ModelScope (see below)  

## Why ModelScope?  

Due to new storage limits introduced by HuggingFace, the GGUF files (30 × 46GB = ~1.38TB) have been moved to ModelScope.  

## Download  

### Python SDK  
```python  
pip install modelscope  
```  

```python  
from modelscope import snapshot_download  
model_dir = snapshot_download('quantzor/DeepSeek-R1-Zero-256x21B-BF16')  
```  

### Direct Link  
🔗 https://modelscope.cn/models/quantzor/DeepSeek-R1-Zero-256x21B-BF16