Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
Datasets:
stanford-crfm
/
DSIR-filtered-pile-50M
like
8
Tasks:
Text Generation
Fill-Mask
Modalities:
Text
Formats:
json
Languages:
English
Size:
1M - 10M
ArXiv:
arxiv:
2302.03169
Tags:
language modeling
masked language modeling
pretraining
pile
DSIR
Croissant
Libraries:
Datasets
Dask
Croissant
License:
mit
Dataset card
Viewer
Files
Files and versions
Community
d0b4e25
DSIR-filtered-pile-50M
1 contributor
History:
13 commits
sangmichaelxie
move all files to train
d0b4e25
10 months ago
.gitattributes
2.71 kB
move all files to train
10 months ago
README.md
4.39 kB
move all files to train
10 months ago
train_1.jsonl
21.3 GB
LFS
add json files
over 1 year ago
train_2.jsonl
21.3 GB
LFS
add json files
over 1 year ago
train_3.jsonl
21.3 GB
LFS
add json files
over 1 year ago
train_4.jsonl
21.1 GB
LFS
add json files
over 1 year ago
train_5.jsonl
83.2 MB
LFS
move all files to train
10 months ago
train_6.jsonl
83.2 MB
LFS
move all files to train
10 months ago