Model-J: MAE Model (model_idx_0121)
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 | 5 |
| Max Train Steps | 1665 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 121 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9614 |
| Val Accuracy | 0.8619 |
| Test Accuracy | 0.8590 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
pickup_truck, man, leopard, bed, lizard, plain, cattle, dolphin, hamster, wardrobe, streetcar, camel, bowl, trout, maple_tree, tiger, palm_tree, fox, lobster, lamp, forest, rocket, tractor, mouse, possum, rose, butterfly, sweet_pepper, motorcycle, crab, castle, shrew, house, clock, tulip, pine_tree, otter, elephant, snake, dinosaur, bear, oak_tree, tank, chair, sea, apple, sunflower, skunk, woman, shark
