Datasets:

oscar-corpus
/

oscar

Tasks:

Text Generation

Sub-tasks:

language-modeling

masked-language-modeling

Languages:

Size:

ArXiv:

License:

Dataset card Files Files and versions Community

Fix `unshuffled_original_ta` split size

#6

by mariosasko - opened Mar 21

base: refs/heads/main

←

from: refs/pr/6

Discussion Files changed

Files changed (1) hide show

README.md +2 -2

README.md CHANGED Viewed

@@ -2529,10 +2529,10 @@ dataset_info:
     dtype: string
   splits:
   - name: train
-    num_bytes: 9934879441
     num_examples: 1263280
   download_size: 1737252172
-  dataset_size: 9934879441
 - config_name: unshuffled_original_tk
   features:
   - name: id

     dtype: string
   splits:
   - name: train
+    num_bytes: 9933590150
     num_examples: 1263280
   download_size: 1737252172
+  dataset_size: 9933590150
 - config_name: unshuffled_original_tk
   features:
   - name: id