metadata
dataset_info:
features:
- name: labels
sequence: int64
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 1087419951
num_examples: 861228
download_size: 376657164
dataset_size: 1087419951