Model-J: ResNet Model (model_idx_0431)
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 | 0.0005 |
| LR Scheduler | constant |
| Epochs | 7 |
| Max Train Steps | 2331 |
| Batch Size | 64 |
| Weight Decay | 0.005 |
| Seed | 431 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|
| Train Accuracy | 0.9523 |
| Val Accuracy | 0.8592 |
| Test Accuracy | 0.8664 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
pear, possum, skyscraper, flatfish, pine_tree, caterpillar, rose, plate, lobster, wardrobe, seal, fox, mushroom, kangaroo, ray, keyboard, aquarium_fish, elephant, television, forest, beetle, orchid, oak_tree, mouse, dinosaur, bowl, snail, cockroach, telephone, hamster, sea, girl, poppy, pickup_truck, streetcar, tiger, crab, chimpanzee, apple, whale, cattle, lion, bee, baby, wolf, worm, rabbit, crocodile, tractor, lizard
