metadata
dataset_info:
features:
- name: text
dtype: string
- name: passage_without_correct_token
dtype: string
- name: original_text
dtype: string
- name: cot
dtype: string
- name: token
dtype: string
splits:
- name: train
num_bytes: 869364
num_examples: 128
download_size: 446310
dataset_size: 869364
configs:
- config_name: default
data_files:
- split: train
path: data/train-*