metadata
base_model: facebook/vit-mae-base
library_name: transformers
pipeline_tag: image-classification
tags:
- probex
- model-j
- weight-space-learning
Model-J: MAE Model (model_idx_0101)
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
🌐 Project | 📃 Paper | 💻 GitHub | 🤗 Dataset
Model Details
| Attribute | Value |
|---|---|
| Subset | MAE |
| Split | val |
| Base Model | facebook/vit-mae-base |
| Dataset | CIFAR100 (50 classes) |
Training Hyperparameters
| Parameter | Value |
|---|---|
| Learning Rate | 0.0005 |
| LR Scheduler | constant_with_warmup |
| Epochs | 8 |
| Max Train Steps | 2664 |
| Batch Size | 64 |
| Weight Decay | 0.005 |
| Seed | 101 |
| Random Crop | True |
| Random Flip | False |
Performance
| Metric | Value |
|---|---|
| Train Accuracy | 0.6843 |
| Val Accuracy | 0.4891 |
| Test Accuracy | 0.5112 |
Training Categories
The model was fine-tuned on the following 50 CIFAR100 classes:
baby, worm, chimpanzee, snail, possum, ray, crab, cattle, table, camel, motorcycle, hamster, wolf, butterfly, trout, mushroom, house, pear, turtle, cloud, castle, beaver, keyboard, poppy, sweet_pepper, clock, lizard, beetle, lion, shark, seal, bottle, aquarium_fish, fox, girl, sea, orange, bee, crocodile, forest, skyscraper, can, road, apple, lamp, plate, rabbit, streetcar, maple_tree, willow_tree