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 | 5e-05 |
| LR Scheduler | constant |
| Epochs | 3 |
| Max Train Steps | 999 |
| Batch Size | 64 |
| Weight Decay | 0.005 |
| Seed | 73 |
| Random Crop | False |
| Random Flip | False |
| Metric | Value |
|---|---|
| Train Accuracy | 0.9808 |
| Val Accuracy | 0.8784 |
| Test Accuracy | 0.8840 |
The model was fine-tuned on the following 50 CIFAR100 classes:
boy, bus, clock, kangaroo, telephone, cloud, hamster, rocket, woman, bottle, road, wardrobe, fox, shark, apple, leopard, skyscraper, man, lobster, dolphin, raccoon, keyboard, pear, turtle, motorcycle, snail, seal, shrew, train, girl, pickup_truck, bear, castle, television, plate, lamp, crocodile, palm_tree, bicycle, cup, worm, dinosaur, spider, pine_tree, sea, plain, rose, camel, possum, trout
Base model
facebook/dino-vitb16