metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
- name: decoder_attention_mask
sequence: int64
splits:
- name: train
num_bytes: 1309686912
num_examples: 196296
- name: test
num_bytes: 10054704
num_examples: 1507
- name: val
num_bytes: 327421728
num_examples: 49074
download_size: 151484595
dataset_size: 1647163344