Model-J: MAE Model (model_idx_0809)
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 | test |
| Base Model | facebook/vit-mae-base |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|
| Learning Rate | 0.0001 |
| LR Scheduler | cosine_with_restarts |
| Epochs | 3 |
| Max Train Steps | 999 |
| Batch Size | 64 |
| Weight Decay | 0.007 |
| Seed | 809 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9625 |
| Val Accuracy | 0.8941 |
| Test Accuracy | 0.8840 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
hamster, streetcar, mountain, bicycle, tiger, castle, mushroom, seal, house, beetle, apple, telephone, pine_tree, lobster, plain, baby, cup, plate, orange, maple_tree, lamp, raccoon, dolphin, bus, caterpillar, cockroach, wardrobe, train, dinosaur, lawn_mower, bed, boy, possum, fox, orchid, rabbit, sweet_pepper, squirrel, woman, beaver, bear, kangaroo, lion, tank, forest, table, otter, poppy, pickup_truck, chair
