metadata
dataset_info:
features:
- name: source
dtype: string
- name: file_type
dtype: string
- name: chunk
dtype: string
- name: chunk_id
dtype: string
splits:
- name: train
num_bytes: 10027415
num_examples: 19722
download_size: 2777200
dataset_size: 10027415
configs:
- config_name: default
data_files:
- split: train
path: data/train-*