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 | cosine |
| Epochs | 7 |
| Max Train Steps | 2331 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 844 |
| Random Crop | True |
| Random Flip | True |
| Metric | Value |
|---|---|
| Train Accuracy | 0.9999 |
| Val Accuracy | 0.9320 |
| Test Accuracy | 0.9304 |
The model was fine-tuned on the following 50 CIFAR100 classes:
maple_tree, mountain, fox, elephant, lobster, woman, mushroom, girl, shrew, road, camel, bee, tank, trout, couch, clock, pear, orange, worm, wardrobe, streetcar, telephone, boy, table, forest, turtle, tiger, shark, beaver, palm_tree, butterfly, orchid, rose, rocket, television, kangaroo, cloud, keyboard, bed, chimpanzee, poppy, bowl, leopard, bear, crocodile, bus, baby, lion, rabbit, skunk
Base model
facebook/dino-vitb16