metadata
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: special_tokens_mask
sequence: int8
splits:
- name: train
num_bytes: 24609228
num_examples: 7945
- name: validation
num_bytes: 2730690
num_examples: 883
download_size: 1204901
dataset_size: 27339918
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*