Model-J: MAE Model (model_idx_0432)
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 | 0.0001 |
| LR Scheduler | linear |
| Epochs | 8 |
| Max Train Steps | 2664 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 432 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9989 |
| Val Accuracy | 0.9037 |
| Test Accuracy | 0.8992 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
chimpanzee, house, lobster, lion, beetle, snake, hamster, kangaroo, bee, bear, possum, tiger, maple_tree, lizard, crocodile, rose, bowl, forest, turtle, bed, butterfly, castle, beaver, mushroom, dolphin, flatfish, girl, skyscraper, mouse, orchid, pickup_truck, telephone, leopard, wardrobe, whale, poppy, keyboard, apple, orange, baby, television, bus, tulip, spider, tank, bottle, table, plain, cup, can
