Model-J: DINO Model (model_idx_0305)
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 | train |
| Base Model | facebook/dino-vitb16 |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|
| Learning Rate | 3e-05 |
| LR Scheduler | cosine |
| Epochs | 4 |
| Max Train Steps | 1332 |
| Batch Size | 64 |
| Weight Decay | 0.01 |
| Seed | 305 |
| Random Crop | False |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9992 |
| Val Accuracy | 0.9248 |
| Test Accuracy | 0.9222 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
willow_tree, otter, tank, caterpillar, keyboard, elephant, man, possum, cup, bowl, lizard, leopard, train, poppy, orange, wardrobe, chimpanzee, forest, plate, butterfly, streetcar, pine_tree, wolf, beaver, skyscraper, woman, rabbit, trout, crab, orchid, clock, dolphin, baby, castle, cattle, sunflower, lobster, oak_tree, whale, crocodile, tiger, road, lamp, mountain, seal, ray, apple, flatfish, worm, cloud
