Model-J: SupViT Model (model_idx_0421)
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 | 9e-05 |
| LR Scheduler | cosine |
| Epochs | 9 |
| Max Train Steps | 2997 |
| Batch Size | 64 |
| Weight Decay | 0.05 |
| Seed | 421 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9996 |
| Val Accuracy | 0.9488 |
| Test Accuracy | 0.9436 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
lobster, chimpanzee, beetle, dinosaur, elephant, tiger, orange, keyboard, man, apple, rabbit, otter, sweet_pepper, castle, lion, beaver, couch, chair, kangaroo, boy, bottle, bus, plate, crocodile, butterfly, maple_tree, raccoon, crab, shrew, pickup_truck, baby, ray, road, squirrel, aquarium_fish, porcupine, television, train, worm, table, sunflower, tulip, bear, bowl, lawn_mower, shark, hamster, tractor, seal, dolphin
