Model-J: MAE Model (model_idx_0503)
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
ProbeXModel 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 | constant_with_warmup |
| Epochs | 4 |
| Max Train Steps | 1332 |
| Batch Size | 64 |
| Weight Decay | 0.007 |
| Seed | 503 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9645 |
| Val Accuracy | 0.8581 |
| Test Accuracy | 0.8542 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
palm_tree, oak_tree, wardrobe, streetcar, lion, beetle, leopard, mouse, otter, shrew, train, possum, house, pine_tree, tank, man, rose, lamp, road, squirrel, can, crab, aquarium_fish, bear, butterfly, chimpanzee, cloud, bottle, forest, crocodile, woman, camel, dolphin, boy, sunflower, tractor, orchid, wolf, spider, caterpillar, shark, kangaroo, tiger, castle, girl, maple_tree, sweet_pepper, plain, bicycle, bowl
