metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 5983337672
num_examples: 2061501
- name: validation
num_bytes: 2782552
num_examples: 985
- name: test
num_bytes: 58812180
num_examples: 20319
download_size: 3003905373
dataset_size: 6044932404
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*