metadata
dataset_info:
features:
- name: bert_token
sequence: int64
- name: gpt2_token
sequence: int64
splits:
- name: train
num_bytes: 830243434.9599016
num_examples: 1207880
- name: test
num_bytes: 667680.9386666666
num_examples: 1156
download_size: 98135244
dataset_size: 830911115.8985683