metadata
dataset_info:
features:
- name: caption
dtype: string
- name: image
dtype: image
- name: link
dtype: string
- name: message_id
dtype: string
- name: timestamp
dtype: string
- name: dense_caption_1
dtype: string
- name: dense_caption_2
dtype: string
- name: dense_caption_3
dtype: string
- name: dense_caption_4
dtype: string
- name: dense_caption_5
dtype: string
- name: dense_caption_6
dtype: string
- name: dense_caption_7
dtype: string
- name: dense_caption_8
dtype: string
- name: dense_caption_9
dtype: string
- name: dense_caption_10
dtype: string
splits:
- name: train
num_bytes: 7529944312.638
num_examples: 4806
download_size: 7512650231
dataset_size: 7529944312.638
configs:
- config_name: default
data_files:
- split: train
path: data/train-*