metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 9403385615
num_examples: 24759486
- name: val
num_bytes: 957032
num_examples: 2500
- name: test
num_bytes: 919749
num_examples: 2500
download_size: 2899144100
dataset_size: 9405262396