metadata
dataset_info:
features:
- name: domain_label
dtype: int64
- name: pass_label
dtype: int64
- name: input
dtype: string
- name: input_ids
sequence: int32
- name: attention_mask
sequence: int8
splits:
- name: train
num_bytes: 431400213
num_examples: 339210
download_size: 123139414
dataset_size: 431400213
configs:
- config_name: default
data_files:
- split: train
path: data/train-*