Hugging Face
Models
Datasets
Spaces
Posts
Docs
Solutions
Pricing
Log In
Sign Up
Datasets:
stanford-crfm
/
DSIR-filtered-pile-50M
like
8
Follow
Stanford CRFM
58
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
Libraries:
Datasets
Dask
Croissant
+ 1
License:
mit
Dataset card
Viewer
Files
Files and versions
Community
bcd7fc6
DSIR-filtered-pile-50M
/
default
1 contributor
History:
3 commits
parquet-converter
Update duckdb index files
bcd7fc6
about 1 year ago
partial-test
Update duckdb index files
about 1 year ago
partial-train
Update parquet files
about 1 year ago
partial-validation
Update duckdb index files
about 1 year ago