Model-J: MAE Model (model_idx_0464)
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 | val |
| Base Model | facebook/vit-mae-base |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|
| Learning Rate | 9e-05 |
| LR Scheduler | constant |
| Epochs | 6 |
| Max Train Steps | 1998 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 464 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9795 |
| Val Accuracy | 0.8640 |
| Test Accuracy | 0.8582 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
bottle, spider, wardrobe, hamster, streetcar, skyscraper, bridge, turtle, fox, couch, poppy, dolphin, tiger, bee, tractor, shrew, willow_tree, dinosaur, whale, bowl, lawn_mower, kangaroo, maple_tree, lamp, table, mouse, orchid, keyboard, crab, pine_tree, road, tank, palm_tree, ray, leopard, flatfish, aquarium_fish, train, chimpanzee, tulip, bicycle, cup, girl, television, pickup_truck, rabbit, rocket, sweet_pepper, can, wolf
