metadata
dataset_info:
features:
- name: Name
dtype: string
- name: cleaned
dtype: string
- name: raw
dtype: string
splits:
- name: train
num_bytes: 450942639
num_examples: 24752
download_size: 291481170
dataset_size: 450942639
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
Read the documentation for the dataset in the [announcement] and in our paper.