Datasets:

Formats:
parquet
Languages:
Chinese
ArXiv:
Libraries:
Datasets
Dask
License:
wanng's picture
Update README.md
ae38067
|
raw
history blame
583 Bytes
metadata
annotations_creators:
  - crowdsourced
language_creators:
  - crowdsourced
language:
  - zh
license:
  - cc-by-4.0
multilinguality:
  - monolingual
pretty_name: laion2B-multi-chinese-subset
size_categories:
  - 10M<n<100M
task_categories:
  - feature-extraction

laion2B-multi-chinese-subset

Wanng: 正在更新。预计国庆后正式上线。

预计一共143M个中文图文对。

Dataset Description

Lisence

CC-BY-4.0