Model-J: MAE Model (model_idx_0425)
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 | 0.0003 |
| LR Scheduler | linear |
| Epochs | 2 |
| Max Train Steps | 666 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 425 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.6918 |
| Val Accuracy | 0.6168 |
| Test Accuracy | 0.6172 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
worm, dolphin, butterfly, hamster, can, bear, skunk, train, raccoon, orange, skyscraper, pear, fox, sweet_pepper, beetle, flatfish, kangaroo, motorcycle, shrew, snake, telephone, willow_tree, rocket, cattle, television, otter, chair, road, forest, house, porcupine, lizard, man, poppy, bed, lion, sea, oak_tree, spider, mouse, tulip, maple_tree, whale, rose, snail, cloud, orchid, shark, clock, bicycle
