Hugging Face
Models
Datasets
Spaces
Posts
Docs
Enterprise
Pricing
Log In
Sign Up
Datasets:
eaglewatch
/
Korean_Wikipedia_Dataset_for_GPT2_August_2022
like
6
Tasks:
Question Answering
Text2Text Generation
Translation
+ 1
Modalities:
Text
Formats:
parquet
Sub-tasks:
open-domain-qa
closed-domain-qa
dialogue-generation
+ 1
Languages:
Korean
Size:
100K - 1M
Tags:
gpt2
korean
wikipedia
pertained
Libraries:
Datasets
Dask
Croissant
+ 1
License:
apache-2.0
Dataset card
Viewer
Files
Files and versions
Community
1
bc0d8bf
Korean_Wikipedia_Dataset_for_GPT2_August_2022
1 contributor
History:
5 commits
eaglewatch
Upload data/valid-00000-of-00001-f75fd8b29148a164.parquet with git-lfs
bc0d8bf
over 1 year ago
data
Upload data/valid-00000-of-00001-f75fd8b29148a164.parquet with git-lfs
over 1 year ago
.gitattributes
Safe
2.31 kB
initial commit
over 1 year ago