metadata
dataset_info:
features:
- name: bert_token
sequence: int64
- name: gpt2_token
sequence: int64
splits:
- name: train
num_bytes: 3998200
num_examples: 20000
- name: test
num_bytes: 199720
num_examples: 900
download_size: 956714
dataset_size: 4197920