metadata
dataset_info:
features:
- name: image
dtype: image
- name: label
dtype: string
- name: original_file_name
dtype: string
splits:
- name: train
num_bytes: 4599929384.25
num_examples: 86030
- name: validation
num_bytes: 606319596.75
num_examples: 10970
- name: test
num_bytes: 565880459
num_examples: 10720
download_size: 5765999984
dataset_size: 5772129440
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: validation
path: data/validation-*
- split: test
path: data/test-*