Hugging Face
Models
Datasets
Spaces
Posts
Docs
Enterprise
Pricing
Log In
Sign Up
Datasets:
wikimedia
/
wit_base
like
56
Follow
Wikimedia
80
Tasks:
Image-to-Text
Text Retrieval
Modalities:
Image
Text
Formats:
parquet
Sub-tasks:
image-captioning
Languages:
Afrikaans
Aragonese
Arabic
+ 102
Size:
100K - 1M
ArXiv:
arxiv:
2103.01913
arxiv:
1512.03385
arxiv:
1905.00641
Tags:
text-image-retrieval
Libraries:
Datasets
Dask
Croissant
+ 1
License:
cc-by-sa-4.0
Dataset card
Viewer
Files
Files and versions
Community
5
refs/convert/parquet
wit_base
/
default
/
partial-train
Commit History
Delete old duckdb index files
fe063a1
verified
parquet-converter
commited on
Mar 8
Update duckdb index files
74982b5
parquet-converter
commited on
Nov 11, 2023
Update parquet files
b9e5286
parquet-converter
commited on
Sep 7, 2023
Update parquet files
53a04f3
parquet-converter
commited on
Aug 21, 2023