shibing624 commited on
Commit
031baba
1 Parent(s): 4443042

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +16 -5
README.md CHANGED
@@ -19,17 +19,28 @@ Chinese Spelling Correction (CSC) is a task to detect and correct misspelled cha
19
 
20
  CSC is challenging since many Chinese characters are visually or phonologically similar but with quite different semantic meanings.
21
 
22
- SIGHAN+Wang271K中文纠错数据集(27万条),是通过原始SIGHAN13、14、15年数据集和Wang271K数据集格式转化后得到,json格式,带错误字符位置信息,SIGHAN为test.json。
23
 
24
 
25
  ### Original Dataset Summary
26
 
27
- - test.json 为 **SIGHAN数据集**, 包括SIGHAN13 14 15,来自 [官方csc.html](http://nlp.ee.ncu.edu.tw/resource/csc.html) ,文件大小:339K,4千条。
28
- - train.json 和 dev.json 为 **Wang271K数据集**,包括 Wang271K ,来自 [Automatic-Corpus-Generation dimmywang提供](https://github.com/wdimmy/Automatic-Corpus-Generation/blob/master/corpus/train.sgml) ,文件大小:93M,27万条。
 
 
 
 
 
 
 
 
 
 
 
29
 
30
  ### Supported Tasks and Leaderboards
31
 
32
- The Alpaca dataset designed for instruction training pretrained language models.
33
 
34
  ### Languages
35
 
@@ -67,7 +78,7 @@ An example of "train" looks as follows:
67
 
68
  | | train | dev | test |
69
  |---------------|------:|--:|--:|
70
- | CSC | 20万条 | 7万条 | 4千条 |
71
 
72
  ## Dataset Creation
73
 
 
19
 
20
  CSC is challenging since many Chinese characters are visually or phonologically similar but with quite different semantic meanings.
21
 
22
+ SIGHAN+Wang271K中文纠错数据集(27万条),是通过原始SIGHAN13、14、15年数据集和Wang271K数据集格式转化后得到,json格式,带错误字符位置信息。
23
 
24
 
25
  ### Original Dataset Summary
26
 
27
+ - test.json 和 dev.json 为 **SIGHAN数据集**, 包括SIGHAN13 14 15,来自 [官方csc.html](http://nlp.ee.ncu.edu.tw/resource/csc.html) ,文件大小:339K,4千条。
28
+ - train.json 为 **Wang271K数据集**,包括 Wang271K ,来自 [Automatic-Corpus-Generation dimmywang提供](https://github.com/wdimmy/Automatic-Corpus-Generation/blob/master/corpus/train.sgml) ,文件大小:93M,27万条。
29
+
30
+ 如果只想用SIGHAN数据集,可以这样取数据:
31
+ ```python
32
+ from datasets import load_dataset
33
+ dev_ds = load_dataset('shibing624/CSC', split='validation')
34
+ print(dev_ds)
35
+ print(dev_ds[0])
36
+ test_ds = load_dataset('shibing624/CSC', split='test')
37
+ print(test_ds)
38
+ print(test_ds[0])
39
+ ```
40
 
41
  ### Supported Tasks and Leaderboards
42
 
43
+ The dataset designed for csc task training pretrained language models.
44
 
45
  ### Languages
46
 
 
78
 
79
  | | train | dev | test |
80
  |---------------|------:|--:|--:|
81
+ | CSC | 251835条 | 27981条 | 1100条 |
82
 
83
  ## Dataset Creation
84