ProbeX/Model-J__DINO__model_idx_0658

Model-J: DINO Model (model_idx_0658)

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	val
Base Model	facebook/dino-vitb16
Dataset	CIFAR100 (50 classes)

Training Hyperparameters

Parameter	Value
Learning Rate	0.0003
LR Scheduler	linear
Epochs	3
Max Train Steps	999
Batch Size	64
Weight Decay	0.03
Seed	658
Random Crop	False
Random Flip	False

Performance

Metric	Value
Train Accuracy	0.4512
Val Accuracy	0.3635
Test Accuracy	0.3848

Training Categories

The model was fine-tuned on the following 50 CIFAR100 classes:

boy, forest, elephant, tiger, lizard, lobster, bowl, worm, pear, seal, couch, trout, beetle, streetcar, wolf, rabbit, clock, kangaroo, oak_tree, tulip, possum, cattle, porcupine, hamster, sunflower, keyboard, orchid, shark, skyscraper, road, telephone, sweet_pepper, raccoon, mouse, tractor, television, cloud, bear, beaver, poppy, man, ray, aquarium_fish, mushroom, rocket, whale, otter, rose, girl, fox