metadata
dataset_info:
features:
- name: date
dtype: string
- name: size
dtype: int64
- name: parsed_date
dtype: timestamp[us]
splits:
- name: train
num_bytes: 604773368
num_examples: 20294152
download_size: 421531443
dataset_size: 604773368
configs:
- config_name: default
data_files:
- split: train
path: data/train-*