Model-J: ResNet Model (model_idx_0758)
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 | linear |
| Epochs | 3 |
| Max Train Steps | 999 |
| Batch Size | 64 |
| Weight Decay | 0.03 |
| Seed | 758 |
| Random Crop | False |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9823 |
| Val Accuracy | 0.9027 |
| Test Accuracy | 0.9020 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
road, bottle, boy, sunflower, raccoon, girl, woman, bear, sweet_pepper, lizard, rose, turtle, shark, lobster, oak_tree, aquarium_fish, plain, dolphin, bus, hamster, tiger, dinosaur, tulip, skyscraper, cup, can, palm_tree, camel, bed, mountain, streetcar, plate, train, shrew, ray, spider, keyboard, castle, cockroach, trout, clock, couch, snake, lamp, orange, crab, leopard, bicycle, beetle, motorcycle
