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 | test |
| Base Model | facebook/dino-vitb16 |
| Dataset | CIFAR100 (50 classes) |
| Parameter | Value |
|---|---|
| Learning Rate | 3e-05 |
| LR Scheduler | cosine |
| Epochs | 9 |
| Max Train Steps | 2997 |
| Batch Size | 64 |
| Weight Decay | 0.01 |
| Seed | 977 |
| Random Crop | False |
| Random Flip | True |
| Metric | Value |
|---|---|
| Train Accuracy | 0.9999 |
| Val Accuracy | 0.9384 |
| Test Accuracy | 0.9366 |
The model was fine-tuned on the following 50 CIFAR100 classes:
pear, sweet_pepper, television, cloud, trout, rocket, plain, bottle, girl, forest, porcupine, streetcar, lion, cattle, snake, bowl, couch, hamster, keyboard, skunk, mouse, shark, plate, leopard, bridge, bee, kangaroo, beetle, butterfly, cup, bicycle, orchid, chimpanzee, bear, aquarium_fish, oak_tree, tractor, poppy, lamp, lobster, raccoon, train, crab, tulip, seal, apple, cockroach, rose, tank, worm
Base model
facebook/dino-vitb16