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 | 0.0001 |
| LR Scheduler | cosine |
| Epochs | 5 |
| Max Train Steps | 1665 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 348 |
| Random Crop | False |
| Random Flip | False |
| Metric | Value |
|---|---|
| Train Accuracy | 0.9996 |
| Val Accuracy | 0.9067 |
| Test Accuracy | 0.9104 |
The model was fine-tuned on the following 50 CIFAR100 classes:
bottle, wolf, shrew, trout, mountain, worm, boy, crab, woman, seal, train, lamp, leopard, mushroom, bus, rabbit, beaver, bear, ray, bowl, butterfly, dinosaur, oak_tree, rose, road, porcupine, sea, table, house, whale, sweet_pepper, keyboard, flatfish, spider, bicycle, couch, forest, bridge, willow_tree, raccoon, skyscraper, camel, orange, orchid, dolphin, clock, lizard, possum, elephant, otter
Base model
facebook/dino-vitb16