metadata
dataset_info:
features:
- name: input_ids
sequence: int32
splits:
- name: train
num_bytes: 5965671696
num_examples: 727876
- name: validation
num_bytes: 2770248
num_examples: 338
- name: test
num_bytes: 58625988
num_examples: 7153
download_size: 2464819303
dataset_size: 6027067932
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*