ProbeX/Model-J__SupViT__model_idx_0009

Model-J: SupViT 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	SupViT
Split	train
Base Model	google/vit-base-patch16-224
Dataset	CIFAR100 (50 classes)

Training Hyperparameters

Parameter	Value
Learning Rate	0.0003
LR Scheduler	cosine_with_restarts
Epochs	7
Max Train Steps	2331
Batch Size	64
Weight Decay	0.009
Seed	9
Random Crop	False
Random Flip	False

Performance

Metric	Value
Train Accuracy	0.9997
Val Accuracy	0.9219
Test Accuracy	0.9228

Training Categories

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

wardrobe, boy, bus, possum, rose, television, lawn_mower, trout, crocodile, hamster, porcupine, whale, lion, apple, pickup_truck, can, palm_tree, bridge, table, maple_tree, cup, chair, beaver, snail, castle, girl, rabbit, orchid, otter, snake, skyscraper, dolphin, cloud, mountain, lizard, spider, aquarium_fish, plate, house, cattle, beetle, seal, orange, worm, flatfish, pear, elephant, mushroom, baby, bowl