Datasets:
metadata
configs:
- config_name: 20k
data_files:
- split: train
path: train/*.tar
- split: test
path: test/*.tar
language:
- en
configs:
- config_name: 20k
data_files:
- split: train
path: train/*.tar
- split: test
path: test/*.tar
language:
- en