Model-J: MAE Model (model_idx_0908)
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 | 5e-05 |
| LR Scheduler | constant_with_warmup |
| Epochs | 7 |
| Max Train Steps | 2331 |
| Batch Size | 64 |
| Weight Decay | 0.005 |
| Seed | 908 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9840 |
| Val Accuracy | 0.8867 |
| Test Accuracy | 0.8662 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
ray, bed, telephone, bicycle, rocket, raccoon, streetcar, sweet_pepper, porcupine, lizard, orange, leopard, elephant, shark, skunk, mushroom, woman, cockroach, bee, chair, rabbit, lobster, cloud, turtle, camel, tulip, whale, pear, lamp, bottle, palm_tree, dolphin, willow_tree, sea, bridge, fox, clock, bear, keyboard, orchid, caterpillar, spider, plain, couch, butterfly, forest, kangaroo, bus, motorcycle, squirrel
