metadata
dataset_info:
features:
- name: text
struct:
- name: attention_mask
sequence: int64
- name: input_ids
sequence: int64
splits:
- name: train
num_bytes: 7567697960
num_examples: 2119719
- name: validation
num_bytes: 76086656
num_examples: 21990
download_size: 816196858
dataset_size: 7643784616