metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 699990424
num_examples: 51334
- name: validation
num_bytes: 69734504
num_examples: 5114
- name: test
num_bytes: 74438924
num_examples: 5459
download_size: 265249928
dataset_size: 844163852
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*