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 | cosine |
| Epochs | 7 |
| Max Train Steps | 2331 |
| Batch Size | 64 |
| Weight Decay | 0.01 |
| Seed | 684 |
| Random Crop | False |
| Random Flip | True |
| Metric | Value |
|---|---|
| Train Accuracy | 0.9998 |
| Val Accuracy | 0.9379 |
| Test Accuracy | 0.9344 |
The model was fine-tuned on the following 50 CIFAR100 classes:
mountain, television, tiger, poppy, road, squirrel, snail, table, man, possum, cup, flatfish, orange, dolphin, orchid, apple, plate, mouse, bicycle, elephant, pickup_truck, skunk, raccoon, maple_tree, worm, tulip, mushroom, snake, cockroach, train, skyscraper, streetcar, leopard, rabbit, lion, lizard, bee, sweet_pepper, beetle, shrew, bear, whale, can, plain, keyboard, wolf, clock, bowl, cloud, lamp
Base model
facebook/dino-vitb16