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 | 0.0003 |
| LR Scheduler | constant_with_warmup |
| Epochs | 3 |
| Max Train Steps | 999 |
| Batch Size | 64 |
| Weight Decay | 0.05 |
| Seed | 662 |
| Random Crop | False |
| Random Flip | False |
| Metric | Value |
|---|---|
| Train Accuracy | 0.4268 |
| Val Accuracy | 0.3627 |
| Test Accuracy | 0.3804 |
The model was fine-tuned on the following 50 CIFAR100 classes:
tulip, pine_tree, beetle, rabbit, skunk, sea, bee, shrew, shark, trout, snake, girl, road, willow_tree, woman, man, motorcycle, telephone, butterfly, dinosaur, tractor, chair, ray, snail, whale, cattle, forest, keyboard, cloud, apple, cup, dolphin, chimpanzee, tank, skyscraper, wolf, worm, otter, pear, tiger, streetcar, leopard, orchid, house, television, castle, baby, train, mouse, seal
Base model
facebook/dino-vitb16