Model-J: MAE Model (model_idx_0939)
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 | 9e-05 |
| LR Scheduler | constant_with_warmup |
| Epochs | 4 |
| Max Train Steps | 1332 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 939 |
| Random Crop | True |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9168 |
| Val Accuracy | 0.8381 |
| Test Accuracy | 0.8302 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
aquarium_fish, table, shark, plate, snake, tulip, beaver, lobster, bee, cup, wolf, snail, ray, sea, pickup_truck, pine_tree, house, poppy, forest, baby, leopard, fox, lion, orchid, palm_tree, whale, butterfly, bowl, caterpillar, motorcycle, dolphin, plain, spider, lamp, possum, dinosaur, streetcar, mouse, can, hamster, shrew, seal, tractor, elephant, pear, mountain, otter, train, girl, rabbit
