Model-J: MAE Model (model_idx_0263)
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 | 9e-05 |
| LR Scheduler | linear |
| Epochs | 6 |
| Max Train Steps | 1998 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 263 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9964 |
| Val Accuracy | 0.8816 |
| Test Accuracy | 0.8764 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
camel, bear, can, rabbit, seal, poppy, rocket, beetle, cup, woman, bridge, forest, raccoon, wolf, wardrobe, whale, bottle, motorcycle, pine_tree, road, turtle, possum, elephant, tulip, couch, worm, tractor, caterpillar, beaver, chair, cattle, leopard, table, trout, oak_tree, man, lion, mushroom, willow_tree, otter, lamp, crab, castle, flatfish, clock, boy, baby, lizard, keyboard, pear
