metadata
dataset_info:
features:
- name: id
dtype: string
- name: media
list: image
- name: messages
dtype: string
splits:
- name: train
num_bytes: 69266986
num_examples: 2000
download_size: 67760940
dataset_size: 69266986
configs:
- config_name: default
data_files:
- split: train
path: data/train-*