Datasets:

Saibo-creator
/

bookcorpus_deduplicated

Modalities:

Formats:

Size:

ArXiv:

Libraries:

Dataset card Viewer Files Files and versions Community

Saibo commited on Dec 28, 2022

Commit

0bb630b

·

1 Parent(s): f6fcc2a

Update README.md

Files changed (1) hide show

README.md +2 -1

README.md CHANGED Viewed

@@ -21,7 +21,8 @@ Effect of deduplication
 - Num of lines: 38832894 VS 74004228
 - Dataset size: 2.91GB VS 4.63GB
-The dataset has been shuffled, so the adjcent texts are no longer consecutive texts in a same book.
 ## Why deduplicate?
 Deduplication of training data has showed various advantages, including:

 - Num of lines: 38832894 VS 74004228
 - Dataset size: 2.91GB VS 4.63GB
+The duplicate text has been droped and only the first appearance is kept.
+The order of text appearance is kept.
 ## Why deduplicate?
 Deduplication of training data has showed various advantages, including: