Model-J: SupViT Model (model_idx_0903)
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 | 0.0003 |
| LR Scheduler | constant_with_warmup |
| Epochs | 3 |
| Max Train Steps | 999 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 903 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9328 |
| Val Accuracy | 0.8515 |
| Test Accuracy | 0.8574 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
snake, flatfish, trout, bottle, willow_tree, mountain, pickup_truck, lizard, oak_tree, seal, baby, raccoon, mushroom, shark, crocodile, possum, poppy, table, worm, boy, maple_tree, tulip, chair, lion, bridge, skunk, television, pine_tree, wolf, rose, tank, hamster, woman, snail, elephant, cockroach, cattle, motorcycle, palm_tree, plain, turtle, aquarium_fish, fox, shrew, lobster, dinosaur, streetcar, bear, otter, couch
