metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: val
path: data/val-*
dataset_info:
features:
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
- name: th-input_ids
sequence: int64
- name: th-attention_mask
sequence: int64
splits:
- name: train
num_bytes: 1087578
num_examples: 666
- name: val
num_bytes: 272711
num_examples: 167
download_size: 31882
dataset_size: 1360289