metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype: string
- name: original_file_name
dtype: string
- name: image_embeddings
sequence: float32
splits:
- name: train
num_bytes: 669486005.5
num_examples: 73660
- name: validation
num_bytes: 91675399.125
num_examples: 9975
- name: test
num_bytes: 91745262.375
num_examples: 10165
download_size: 986967567
dataset_size: 852906667
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*