ProbeX/Model-J__MAE__model_idx_0003

Model-J: MAE Model (model_idx_0003)

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	MAE
Split	train
Base Model	facebook/vit-mae-base
Dataset	CIFAR100 (50 classes)

Training Hyperparameters

Parameter	Value
Learning Rate	7e-05
LR Scheduler	cosine_with_restarts
Epochs	3
Max Train Steps	999
Batch Size	64
Weight Decay	0.009
Seed	3
Random Crop	True
Random Flip	False

Performance

Metric	Value
Train Accuracy	0.9438
Val Accuracy	0.8632
Test Accuracy	0.8670

Training Categories

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

camel, rabbit, clock, forest, pear, cloud, shrew, raccoon, shark, snail, bus, leopard, cockroach, boy, turtle, kangaroo, crab, snake, elephant, caterpillar, pine_tree, worm, bed, dolphin, spider, can, road, baby, cup, house, oak_tree, beaver, wolf, sea, table, beetle, lamp, flatfish, porcupine, whale, tiger, skyscraper, willow_tree, rocket, skunk, bear, orchid, bowl, fox, wardrobe