ProbeX/Model-J__DINO__model_idx_0009

Model-J: DINO Model (model_idx_0009)

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

Training Hyperparameters

Parameter	Value
Learning Rate	0.0005
LR Scheduler	cosine_with_restarts
Epochs	3
Max Train Steps	999
Batch Size	64
Weight Decay	0.03
Seed	9
Random Crop	False
Random Flip	True

Performance

Metric	Value
Train Accuracy	0.3387
Val Accuracy	0.3037
Test Accuracy	0.3170

Training Categories

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

cattle, squirrel, sweet_pepper, pine_tree, lizard, spider, snake, sunflower, tulip, caterpillar, snail, rose, cloud, tractor, clock, pear, raccoon, otter, porcupine, rocket, shrew, cup, leopard, motorcycle, bridge, man, sea, baby, bee, road, tank, worm, oak_tree, lamp, camel, cockroach, hamster, lobster, woman, orange, shark, ray, can, beaver, willow_tree, skunk, beetle, trout, crab, apple