metadata
dataset_info:
features:
- name: input_ids
sequence: int64
- name: labels
sequence: int64
- name: Direction
dtype: string
- name: __index_level_0__
dtype: int64
splits:
- name: train
num_bytes: 59446759
num_examples: 86448
download_size: 16655555
dataset_size: 59446759
configs:
- config_name: default
data_files:
- split: train
path: data/train-*