Model-J: SupViT Model (model_idx_0535)
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 | val |
| Base Model | google/vit-base-patch16-224 |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|
| Learning Rate | 0.0003 |
| LR Scheduler | constant |
| Epochs | 5 |
| Max Train Steps | 1665 |
| Batch Size | 64 |
| Weight Decay | 0.005 |
| Seed | 535 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9237 |
| Val Accuracy | 0.8723 |
| Test Accuracy | 0.8730 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
orange, train, butterfly, whale, elephant, leopard, man, spider, snail, beetle, lizard, seal, television, pine_tree, pickup_truck, plain, couch, orchid, bed, rose, wolf, worm, rabbit, skunk, bee, baby, crab, snake, mushroom, mountain, streetcar, trout, skyscraper, clock, tank, poppy, lobster, can, tractor, bear, cup, girl, forest, possum, woman, sweet_pepper, squirrel, palm_tree, aquarium_fish, shrew
