metadata
base_model: facebook/dino-vitb16
library_name: transformers
pipeline_tag: image-classification
tags:
- probex
- model-j
- weight-space-learning
Model-J: DINO Model (model_idx_0012)
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
🌐 Project | 📃 Paper | 💻 GitHub | 🤗 Dataset
Model Details
| Attribute | Value |
|---|---|
| Subset | DINO |
| Split | test |
| Base Model | facebook/dino-vitb16 |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|---|
| Learning Rate | 7e-05 |
| LR Scheduler | constant |
| Epochs | 6 |
| Max Train Steps | 1998 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 12 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|---|
| Train Accuracy | 0.9872 |
| Val Accuracy | 0.8891 |
| Test Accuracy | 0.8828 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
can, woman, mountain, whale, bee, cup, man, motorcycle, forest, plate, kangaroo, oak_tree, sea, bed, bear, crab, shrew, otter, willow_tree, beetle, lobster, tiger, snake, lizard, keyboard, telephone, elephant, castle, house, clock, pear, snail, streetcar, rose, tractor, dinosaur, sweet_pepper, tulip, fox, spider, plain, leopard, orange, bicycle, chimpanzee, chair, flatfish, sunflower, skyscraper, porcupine