Model-J: DINO Model (model_idx_0822)
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 | DINO |
| Split | test |
| Base Model | facebook/dino-vitb16 |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|
| Learning Rate | 0.0001 |
| LR Scheduler | cosine |
| Epochs | 3 |
| Max Train Steps | 999 |
| Batch Size | 64 |
| Weight Decay | 0.007 |
| Seed | 822 |
| Random Crop | True |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9959 |
| Val Accuracy | 0.9205 |
| Test Accuracy | 0.9248 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
bed, aquarium_fish, bear, seal, train, spider, tulip, caterpillar, crab, keyboard, cloud, house, lamp, crocodile, leopard, bee, clock, lawn_mower, bicycle, lobster, kangaroo, wolf, tiger, tractor, beetle, palm_tree, hamster, fox, dinosaur, poppy, plate, sweet_pepper, otter, plain, porcupine, butterfly, man, sunflower, snail, shark, pickup_truck, maple_tree, sea, bridge, squirrel, mountain, girl, table, rabbit, baby
