Model-J: SupViT Model (model_idx_0887)
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 | 0.0001 |
| LR Scheduler | linear |
| Epochs | 4 |
| Max Train Steps | 1332 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 887 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9994 |
| Val Accuracy | 0.9589 |
| Test Accuracy | 0.9600 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
wolf, worm, bicycle, willow_tree, lawn_mower, palm_tree, tulip, bed, cloud, bee, bus, tiger, bottle, bear, woman, lobster, rose, apple, kangaroo, plate, television, ray, chimpanzee, snake, orchid, aquarium_fish, beaver, raccoon, mouse, shark, wardrobe, squirrel, poppy, fox, couch, otter, sunflower, baby, man, table, camel, can, dinosaur, beetle, dolphin, turtle, lizard, tank, pear, castle
