ProbeX/Model-J__DINO__model_idx_0008

Model-J: DINO Model (model_idx_0008)

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	7e-05
LR Scheduler	cosine_with_restarts
Epochs	9
Max Train Steps	2997
Batch Size	64
Weight Decay	0.05
Seed	8
Random Crop	True
Random Flip	False

Performance

Metric	Value
Train Accuracy	0.9998
Val Accuracy	0.9141
Test Accuracy	0.9126

Training Categories

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

oak_tree, worm, skyscraper, snake, rabbit, telephone, streetcar, rocket, possum, table, plain, rose, crocodile, chimpanzee, castle, man, keyboard, couch, aquarium_fish, bicycle, mushroom, otter, cup, wolf, plate, can, orange, sunflower, willow_tree, boy, road, tulip, mouse, forest, bridge, girl, house, cattle, bee, lobster, flatfish, television, bed, crab, pickup_truck, baby, whale, trout, kangaroo, skunk