---
title: stack-exchange-paired
canonical_url: "https://www.modelscope.cn/datasets/AI-ModelScope/stack-exchange-paired"
md_url: "https://www.modelscope.cn/datasets/AI-ModelScope/stack-exchange-paired.md"
repository: AI-ModelScope/stack-exchange-paired
chinese_name: stack-exchange-paired
last_updated: 2024-09-02
license: "Apache License 2.0"
storage_size: "70 GB"
domain:
  - configs
  - size_categories
  - language
  - task_categories
  - pretty_name
tasks:
  - "{config_name=rl, data_files=[{path=data/rl/merged_rl.csv, split=train}]}"
  - "10M<n<100M"
  - en
  - question-answering
  - "StackExchange Paired"
downloads: 727
stars: 2
---

# stack-exchange-paired

> stack-exchange-paired - AI-ModelScope 在 ModelScope 开源的数据集。StackExchange Paired

AI-ModelScope/stack-exchange-paired 是 ModelScope 魔搭社区上的{config_name=rl, data_files=[{path=data/rl/merged_rl.csv, split=train}]}、10M<n<100M、en数据集，涉及 configs、size_categories、language 领域，存储大小 70 GB，采用 Apache License 2.0 许可。

- **Repository**: AI-ModelScope/stack-exchange-paired
- **License**: Apache License 2.0
- **Tasks**: {config_name=rl, data_files=[{path=data/rl/merged_rl.csv, split=train}]}, 10M<n<100M, en, question-answering, StackExchange Paired
- **Domain**: configs, size_categories, language, task_categories, pretty_name
- **Storage size**: 70 GB
- **Downloads**: 727
- **Stars**: 2
- **Last updated**: 2024-09-02

Source: https://www.modelscope.cn/datasets/AI-ModelScope/stack-exchange-paired

---

# StackExchange Paired

This is a processed version of the [`HuggingFaceH4/stack-exchange-preferences`](https://huggingface.co/datasets/HuggingFaceH4/stack-exchange-preferences). The following steps were applied:

- Parse HTML to Markdown with `markdownify`
- Create pairs `(response_j, response_k)` where j was rated better than k
- Sample at most 10 pairs per question
- Shuffle the dataset globally

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

ds = MsDataset.load('AI-ModelScope/stack-exchange-paired',subset_name='default', split='train', download_mode=DownloadMode.FORCE_REDOWNLOAD)  
print(next(iter(ds)))
```

This dataset is designed to be used for preference learning. The processing notebook is in [the repository](https://huggingface.co/datasets/lvwerra/stack-exchange-paired/tree/main) as well.
