Model-J: DINO Model (model_idx_0157)
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 | linear |
| Epochs | 7 |
| Max Train Steps | 2331 |
| Batch Size | 64 |
| Weight Decay | 0.007 |
| Seed | 157 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9992 |
| Val Accuracy | 0.9077 |
| Test Accuracy | 0.8998 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
girl, bus, seal, clock, orchid, mushroom, apple, sunflower, couch, mouse, shark, lawn_mower, pine_tree, plate, tank, sweet_pepper, road, trout, lion, poppy, bee, rose, elephant, wolf, crab, cup, cattle, rocket, ray, kangaroo, chimpanzee, streetcar, chair, squirrel, lobster, oak_tree, boy, beaver, skunk, castle, sea, whale, porcupine, otter, wardrobe, house, fox, possum, can, dolphin
