metadata
dataset_info:
features:
- name: image
dtype: image
- name: id
dtype: int64
- name: label
dtype: float64
- name: text
dtype: string
- name: split
dtype: string
splits:
- name: train
num_bytes: 2983171367
num_examples: 8500
- name: validation
num_bytes: 189052953
num_examples: 500
- name: test
num_bytes: 361452926
num_examples: 1000
download_size: 3605181786
dataset_size: 3533677246
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*