Model-J: MAE Model (model_idx_0629)
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 | 3e-05 |
| LR Scheduler | linear |
| Epochs | 5 |
| Max Train Steps | 1665 |
| Batch Size | 64 |
| Weight Decay | 0.007 |
| Seed | 629 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9795 |
| Val Accuracy | 0.8592 |
| Test Accuracy | 0.8612 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
apple, motorcycle, kangaroo, table, caterpillar, dinosaur, chimpanzee, mushroom, leopard, lobster, butterfly, pine_tree, lion, ray, rocket, lamp, beaver, seal, wardrobe, can, porcupine, baby, girl, streetcar, tank, mouse, dolphin, bus, telephone, lizard, cattle, skyscraper, plain, raccoon, chair, boy, turtle, maple_tree, bridge, orchid, whale, rose, otter, pickup_truck, sunflower, bicycle, trout, shrew, flatfish, couch
