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: 58030544
num_examples: 22480
download_size: 16743699
dataset_size: 58030544
configs:
- config_name: default
data_files:
- split: train
path: data/train-*