metadata
dataset_info:
features:
- name: id
dtype: int32
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 15697219153
num_examples: 20480000
- name: val
num_bytes: 3827042
num_examples: 5000
- name: test
num_bytes: 7670994
num_examples: 10000
download_size: 7817630008
dataset_size: 15708717189