Model-J: SupViT Model (model_idx_0605)
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 | test |
| Base Model | google/vit-base-patch16-224 |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|
| Learning Rate | 0.0005 |
| LR Scheduler | constant |
| Epochs | 8 |
| Max Train Steps | 2664 |
| Batch Size | 64 |
| Weight Decay | 0.005 |
| Seed | 605 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9602 |
| Val Accuracy | 0.8589 |
| Test Accuracy | 0.8488 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
bicycle, caterpillar, leopard, dinosaur, kangaroo, flatfish, fox, bed, house, lawn_mower, bear, television, rocket, palm_tree, train, cattle, skyscraper, tank, road, lamp, cup, worm, lizard, chimpanzee, orange, bottle, whale, pickup_truck, tiger, cockroach, chair, poppy, motorcycle, bus, can, rose, girl, oak_tree, keyboard, mouse, snake, streetcar, sea, sweet_pepper, elephant, telephone, pine_tree, aquarium_fish, lion, mushroom
