ProbeX/Model-J__SupViT__model_idx_0005

Model-J: SupViT Model (model_idx_0005)

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	val
Base Model	google/vit-base-patch16-224
Dataset	CIFAR100 (50 classes)

Training Hyperparameters

Parameter	Value
Learning Rate	7e-05
LR Scheduler	cosine
Epochs	9
Max Train Steps	2997
Batch Size	64
Weight Decay	0.03
Seed	5
Random Crop	True
Random Flip	True

Performance

Metric	Value
Train Accuracy	0.9994
Val Accuracy	0.9421
Test Accuracy	0.9420

Training Categories

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

tulip, shrew, flatfish, crocodile, bus, streetcar, tractor, otter, tank, caterpillar, house, bee, pear, willow_tree, bridge, dinosaur, butterfly, cattle, can, spider, crab, bicycle, sea, apple, bed, snake, oak_tree, porcupine, worm, squirrel, orange, skyscraper, rabbit, raccoon, seal, camel, plain, sweet_pepper, beaver, girl, elephant, snail, orchid, keyboard, cloud, boy, ray, bottle, rose, tiger