Model-J: DINO Model (model_idx_0239)
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 | 0.0003 |
| LR Scheduler | linear |
| Epochs | 7 |
| Max Train Steps | 2331 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 239 |
| Random Crop | True |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.5794 |
| Val Accuracy | 0.4432 |
| Test Accuracy | 0.4420 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
kangaroo, mountain, castle, bee, raccoon, turtle, television, cattle, streetcar, bear, caterpillar, bed, telephone, beetle, plate, girl, cup, couch, ray, rocket, otter, man, tractor, chair, worm, porcupine, baby, fox, orange, house, clock, skyscraper, possum, sweet_pepper, palm_tree, chimpanzee, motorcycle, shark, boy, bridge, pickup_truck, bus, maple_tree, trout, willow_tree, can, lamp, pine_tree, crocodile, cockroach
