metadata
dataset_info:
features:
- name: category
dtype: string
- name: articles
sequence: string
- name: num_tokens
dtype: int64
- name: chunks
list:
- name: chunk
dtype: string
- name: chunk_size
dtype: int64
- name: start
dtype: int64
splits:
- name: train
num_bytes: 399676341
num_examples: 184
download_size: 192015010
dataset_size: 399676341
configs:
- config_name: default
data_files:
- split: train
path: data/train-*