Model-J DINO
Collection
1002 items
โข
Updated
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
๐ Project | ๐ Paper | ๐ป GitHub | ๐ค Dataset
| Attribute | Value |
|---|---|
| Subset | DINO |
| Split | train |
| Base Model | facebook/dino-vitb16 |
| Dataset | CIFAR100 (50 classes) |
| Parameter | Value |
|---|---|
| Learning Rate | 3e-05 |
| LR Scheduler | linear |
| Epochs | 4 |
| Max Train Steps | 1332 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 309 |
| Random Crop | False |
| Random Flip | False |
| Metric | Value |
|---|---|
| Train Accuracy | 0.9996 |
| Val Accuracy | 0.9109 |
| Test Accuracy | 0.9174 |
The model was fine-tuned on the following 50 CIFAR100 classes:
snake, sea, pear, lion, willow_tree, table, rocket, shrew, fox, ray, man, porcupine, plate, whale, raccoon, palm_tree, bowl, motorcycle, tulip, flatfish, squirrel, oak_tree, bicycle, elephant, camel, orchid, hamster, streetcar, plain, tractor, bridge, pine_tree, aquarium_fish, can, rabbit, caterpillar, baby, bear, beetle, otter, sunflower, bee, worm, boy, road, house, leopard, mouse, possum, dolphin
Base model
facebook/dino-vitb16