metadata
dataset_info:
features:
- name: doc_id
dtype: large_string
- name: text
dtype: large_string
- name: type
dtype: large_string
splits:
- name: train
num_bytes: 2816325279.2
num_examples: 500000
download_size: 1305956338
dataset_size: 2816325279.2
configs:
- config_name: default
data_files:
- split: train
path: data/train-*