Model-J: SupViT Model (model_idx_0102)
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 | 0.0003 |
| LR Scheduler | constant_with_warmup |
| Epochs | 5 |
| Max Train Steps | 1665 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 102 |
| Random Crop | False |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9595 |
| Val Accuracy | 0.8752 |
| Test Accuracy | 0.8766 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
trout, otter, rabbit, plain, lion, mountain, dolphin, train, crab, television, bicycle, hamster, sea, chair, poppy, elephant, turtle, mushroom, boy, butterfly, porcupine, clock, road, tank, baby, rocket, man, skyscraper, motorcycle, rose, bed, spider, beetle, bear, skunk, bee, whale, girl, snail, wardrobe, oak_tree, orchid, orange, pine_tree, bridge, tulip, can, mouse, snake, sweet_pepper
