Model-J: SupViT Model (model_idx_0531)
This model is part of the Model-J dataset, introduced in:
Learning on Model Weights using Tree Experts (CVPR 2025) by Eliahu Horwitz*, Bar Cavia*, Jonathan Kahana*, Yedid Hoshen
ProbeXModel Details
| Attribute | Value |
|---|
| Subset | SupViT |
| Split | train |
| Base Model | google/vit-base-patch16-224 |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|
| Learning Rate | 0.0003 |
| LR Scheduler | constant_with_warmup |
| Epochs | 5 |
| Max Train Steps | 1665 |
| Batch Size | 64 |
| Weight Decay | 0.007 |
| Seed | 531 |
| Random Crop | True |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9561 |
| Val Accuracy | 0.8584 |
| Test Accuracy | 0.8534 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
wolf, plate, beetle, aquarium_fish, bottle, apple, pine_tree, tank, ray, sunflower, chimpanzee, streetcar, mouse, kangaroo, sweet_pepper, porcupine, plain, squirrel, lion, leopard, seal, sea, hamster, rocket, motorcycle, baby, orange, maple_tree, skyscraper, lamp, willow_tree, crocodile, possum, flatfish, palm_tree, dinosaur, bus, camel, castle, fox, caterpillar, bee, girl, snail, beaver, trout, oak_tree, raccoon, bear, lobster
