metadata
dataset_info:
features:
- name: text
dtype: string
- name: source
dtype: string
splits:
- name: train
num_bytes: 3532055082
num_examples: 3000000
- name: validation
num_bytes: 1177629097
num_examples: 1000000
- name: test
num_bytes: 1176985149
num_examples: 1000000
download_size: 3309328042
dataset_size: 5886669328
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*