Model-J: SupViT Model (model_idx_0774)
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 | 3e-05 |
| LR Scheduler | linear |
| Epochs | 3 |
| Max Train Steps | 999 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 774 |
| Random Crop | True |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9794 |
| Val Accuracy | 0.9400 |
| Test Accuracy | 0.9386 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
boy, possum, maple_tree, sea, tractor, sunflower, keyboard, tiger, pear, rose, tulip, wolf, rocket, camel, snail, bicycle, poppy, road, mouse, spider, caterpillar, lizard, skunk, motorcycle, plate, butterfly, lamp, woman, bear, dinosaur, seal, orange, castle, plain, forest, man, wardrobe, pickup_truck, tank, chair, mushroom, worm, crab, lawn_mower, clock, trout, sweet_pepper, willow_tree, crocodile, oak_tree
