Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
Datasets:
xz56
/
openwebtext-tokenized-small
like
0
Formats:
parquet
Size:
1M - 10M
Libraries:
Datasets
Dask
Croissant
+ 1
Dataset card
Viewer
Files
Files and versions
Community
1
xz56
commited on
Mar 29
Commit
55fe8f9
•
1 Parent(s):
b27534b
Update README.md
Browse files
Files changed (1)
hide
show
README.md
+1
-0
README.md
CHANGED
Viewed
@@ -20,3 +20,4 @@ configs:
20
- split: test
21
path: data/test-*
22
---
20
- split: test
21
path: data/test-*
22
---
23
+
The first 15% of Openwebtext2 tokenized with LLamaTokenizer for 256 context length. Total of ~1.5 Billion tokens.