Model-J: SupViT Model (model_idx_0086)
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 | 7e-05 |
| LR Scheduler | linear |
| Epochs | 6 |
| Max Train Steps | 1998 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 86 |
| Random Crop | False |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9998 |
| Val Accuracy | 0.9347 |
| Test Accuracy | 0.9384 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
maple_tree, kangaroo, castle, orange, ray, lion, mouse, girl, sea, spider, train, pear, dinosaur, raccoon, camel, tiger, cloud, lamp, chimpanzee, bridge, bee, boy, pickup_truck, rose, whale, lobster, clock, house, otter, leopard, rabbit, bowl, turtle, caterpillar, man, cockroach, bus, snake, dolphin, streetcar, shark, butterfly, tank, tractor, forest, beetle, willow_tree, couch, oak_tree, hamster
