Model-J: MAE Model (model_idx_0069)
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 | cosine |
| Epochs | 8 |
| Max Train Steps | 2664 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 69 |
| Random Crop | False |
| Random Flip | True |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9991 |
| Val Accuracy | 0.9141 |
| Test Accuracy | 0.9106 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
apple, table, plate, orange, porcupine, wardrobe, couch, castle, train, palm_tree, man, sunflower, tractor, boy, cup, shark, tulip, butterfly, streetcar, tiger, squirrel, crocodile, clock, mountain, mouse, road, bicycle, whale, bus, house, shrew, keyboard, sea, tank, poppy, bridge, bottle, flatfish, raccoon, chair, skunk, oak_tree, kangaroo, bed, cloud, leopard, rabbit, woman, aquarium_fish, worm
