metadata
dataset_info:
features:
- name: docid
dtype: string
- name: text
dtype: string
splits:
- name: train
num_bytes: 15677113
num_examples: 19356
download_size: 9334801
dataset_size: 15677113
configs:
- config_name: default
data_files:
- split: train
path: data/train-*