File size: 2,198 Bytes
a857937
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
---
license: mit
configs:
- config_name: default
  data_files:
  - split: train
    path: data/train-*
dataset_info:
  features:
  - name: _id
    dtype: string
  - name: article
    dtype: string
  - name: question
    dtype: string
  - name: option
    sequence: string
  - name: answer
    dtype: string
  splits:
  - name: train
    num_bytes: 174814003
    num_examples: 97687
  download_size: 41003816
  dataset_size: 174814003
---



RACE 是一个大规模阅读理解数据集,包含超过 28,000 篇文章以及近 100,000 个问题。该数据集采集自中国的中学和高中英语考试,可以作为机器阅读理解任务的训练与测试集。

RACE is a large-scale reading comprehension dataset with over 28,000 passages and nearly 100,000 questions. The dataset is collected from English exams for Chinese middle and high school students and can serve as training and test sets for machine comprehension tasks.

---

### 数据格式 | Data Format

- **article(文章)**: 一个字符串,包含完整的阅读文章。
- **questions(问题)**: 一个字符串列表,每个字符串为一个问题(有陈述句和带填空两类)。
- **option(选项)**: 列表,每个问题有四个备选答案。
- **answer(答案)**: 正确答案的标签(如"A","B","C","D")。
- **_id(唯一编号)**: 每篇文章的唯一标识。

---

### 使用说明 | Notes

1. RACE 数据集仅限非商业科研使用。
2. 所有文章均采集自互联网,不属于 Carnegie Mellon University 所有。
3. 不得以任何商业目的复制、出售、转售、利用数据集内容或衍生数据。
4. 官方有权终止对数据集的访问权限。

---

### 论文引用 | Paper Reference

> RACE: Large-scale ReAding Comprehension Dataset From Examinations  
> Guokun Lai*, Qizhe Xie*, Hanxiao Liu, Yiming Yang and Eduard Hovy  
> arXiv preprint arXiv:1704.04683

---

### 联系方式 | Contact

如有问题,请联系 Guokun Lai 和 Qizhe Xie。

---

### 许可协议 | License

MIT license,仅限研究用途,禁止商业化。

---

如需使用, 请通过[[官方链接](https://www.cs.cmu.edu/~glai1/data/race/)]下载数据集。

---