Model-J: DINO Model (model_idx_0578)
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 | cosine_with_restarts |
| Epochs | 9 |
| Max Train Steps | 2997 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 578 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.8796 |
| Val Accuracy | 0.4992 |
| Test Accuracy | 0.5160 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
bridge, sea, beaver, rose, television, skyscraper, chair, woman, man, lamp, poppy, possum, mouse, bee, girl, ray, bus, squirrel, tractor, tiger, baby, otter, chimpanzee, palm_tree, maple_tree, oak_tree, bicycle, fox, table, orange, kangaroo, crocodile, leopard, cattle, shark, keyboard, rocket, apple, raccoon, forest, wardrobe, cockroach, plain, flatfish, house, orchid, caterpillar, cloud, telephone, road
