---
title: SFT-Nectar
canonical_url: "https://www.modelscope.cn/datasets/AI-ModelScope/SFT-Nectar"
md_url: "https://www.modelscope.cn/datasets/AI-ModelScope/SFT-Nectar.md"
repository: AI-ModelScope/SFT-Nectar
chinese_name: SFT-Nectar
last_updated: 2024-09-02
license: cc-by-nc-4.0
storage_size: "235 MB"
domain:
  - configs
tasks:
  - "{config_name=default, data_files=[{path=sft_data_structured.jsonl, split=train}]}"
downloads: 1536
stars: 0
---

# SFT-Nectar

> SFT-Nectar - AI-ModelScope 在 ModelScope 开源的数据集。

AI-ModelScope/SFT-Nectar 是 ModelScope 魔搭社区上的{config_name=default, data_files=[{path=sft_data_structured.jsonl, split=train}]}数据集，涉及 configs 领域，存储大小 235 MB，采用 cc-by-nc-4.0 许可。

- **Repository**: AI-ModelScope/SFT-Nectar
- **License**: cc-by-nc-4.0
- **Tasks**: {config_name=default, data_files=[{path=sft_data_structured.jsonl, split=train}]}
- **Domain**: configs
- **Storage size**: 235 MB
- **Downloads**: 1536
- **Stars**: 0
- **Last updated**: 2024-09-02

Source: https://www.modelscope.cn/datasets/AI-ModelScope/SFT-Nectar

---

## 示例代码
```python
from modelscope import MsDataset
from modelscope.utils.constant import DownloadMode

ds = MsDataset.load('AI-ModelScope/SFT-Nectar',subset_name='default', split='train', download_mode=DownloadMode.FORCE_REDOWNLOAD, cache_dir='/mnt/workspace/cache_yk')  
print(next(iter(ds)))
```
