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: 2013563824
num_examples: 218107
- name: val
num_bytes: 503393264
num_examples: 54527
download_size: 651358784
dataset_size: 2516957088