metadata
dataset_info:
features:
- name: index
dtype: int64
- name: item_descriptions
dtype: string
- name: item_description_tokens
sequence: int64
splits:
- name: train
num_bytes: 21316632
num_examples: 4819
download_size: 1270187
dataset_size: 21316632
configs:
- config_name: default
data_files:
- split: train
path: data/train-*