Model-J: ResNet Model (model_idx_0310)
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 | ResNet |
| Split | test |
| Base Model | microsoft/resnet-101 |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|
| Learning Rate | 0.0003 |
| LR Scheduler | cosine |
| Epochs | 3 |
| Max Train Steps | 999 |
| Batch Size | 64 |
| Weight Decay | 0.05 |
| Seed | 310 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9542 |
| Val Accuracy | 0.8885 |
| Test Accuracy | 0.8894 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
motorcycle, worm, butterfly, porcupine, dinosaur, cockroach, tulip, orchid, lawn_mower, aquarium_fish, cattle, keyboard, road, forest, spider, sunflower, snake, bridge, orange, poppy, chair, house, rabbit, lamp, bee, seal, plate, bicycle, camel, chimpanzee, shrew, palm_tree, bottle, clock, whale, skyscraper, skunk, bear, baby, shark, lion, caterpillar, mountain, raccoon, squirrel, bowl, possum, streetcar, crocodile, tractor
