ProbeX/Model-J__DINO__model_idx_0521

Model-J: DINO Model (model_idx_0521)

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	0.0001
LR Scheduler	constant
Epochs	8
Max Train Steps	2664
Batch Size	64
Weight Decay	0.005
Seed	521
Random Crop	True
Random Flip	False

Performance

Metric	Value
Train Accuracy	0.9669
Val Accuracy	0.8536
Test Accuracy	0.8478

Training Categories

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

skunk, cloud, bear, tulip, motorcycle, shark, man, bed, mountain, orange, clock, wardrobe, pickup_truck, keyboard, rabbit, forest, apple, sunflower, possum, tiger, whale, house, snake, maple_tree, beetle, chimpanzee, seal, ray, bowl, girl, chair, table, tractor, mushroom, caterpillar, rose, castle, spider, rocket, elephant, couch, boy, hamster, train, skyscraper, lobster, beaver, cockroach, orchid, bicycle