ProbeX/Model-J__MAE__model_idx_0000

Model-J: MAE Model (model_idx_0000)

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	0.0005
LR Scheduler	cosine
Epochs	6
Max Train Steps	1998
Batch Size	64
Weight Decay	0.009
Seed	0
Random Crop	False
Random Flip	False

Performance

Metric	Value
Train Accuracy	0.7001
Val Accuracy	0.4885
Test Accuracy	0.4934

Training Categories

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

skyscraper, rose, squirrel, beetle, telephone, oak_tree, caterpillar, ray, bus, dolphin, girl, bridge, poppy, flatfish, woman, rocket, bear, television, dinosaur, tank, motorcycle, lizard, orchid, cloud, sea, mountain, streetcar, shrew, wardrobe, can, palm_tree, baby, lamp, kangaroo, willow_tree, beaver, possum, aquarium_fish, lobster, bicycle, otter, crocodile, plain, wolf, bowl, cockroach, fox, cup, lawn_mower, skunk