Model-J: MAE Model (model_idx_0048)
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 | 5e-05 |
| LR Scheduler | constant_with_warmup |
| Epochs | 2 |
| Max Train Steps | 666 |
| Batch Size | 64 |
| Weight Decay | 0.005 |
| Seed | 48 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9157 |
| Val Accuracy | 0.8451 |
| Test Accuracy | 0.8522 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
chair, rabbit, train, tiger, fox, lion, willow_tree, raccoon, boy, sunflower, bowl, keyboard, dinosaur, bottle, lobster, caterpillar, snake, lawn_mower, couch, bed, orange, palm_tree, cloud, telephone, streetcar, cockroach, bicycle, kangaroo, motorcycle, poppy, dolphin, beaver, rocket, clock, hamster, apple, wardrobe, bridge, bus, lizard, plate, beetle, chimpanzee, porcupine, crab, tulip, shark, mushroom, baby, aquarium_fish
