ImageNet-1k's validation split (taken from
https://www.kaggle.com/datasets/sautkin/imagenet1kvalid/data) and filtered into the 100 labels corresponding to the following dict for EECS542:
imagenet100_to_name = {0: 'robin', 1: 'Gila_monster', 2: 'hognose_snake', 3: 'garter_snake', 4: 'green_mamba', 5: 'garden_spider', 6: 'lorikeet', 7: 'goose', 8: 'rock_crab', 9: 'fiddler_crab', 10: 'American_lobster', 11: 'little_blue_heron', 12: 'American_coot', 13: 'Chihuahua', 14: 'Shih-Tzu', 15: 'papillon'… See the full description on the dataset page:
https://huggingface.co/datasets/cjschul/eecs542_imagenet_100.
