metadata
dataset_info:
features:
- name: caption
dtype: string
- name: link
dtype: string
- name: message_id
dtype: string
- name: timestamp
dtype: string
splits:
- name: train
num_bytes: 204134
num_examples: 96
download_size: 111233
dataset_size: 204134
configs:
- config_name: default
data_files:
- split: train
path: data/train-*