README.md

---
base_model: microsoft/resnet-101
library_name: transformers
pipeline_tag: image-classification
tags:
  - probex
  - model-j
  - weight-space-learning
---

# Model-J: ResNet Model (model_idx_0433)

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

<p align="center">
    🌐 <a href="https://horwitz.ai/probex" target="_blank">Project</a> | 📃 <a href="https://arxiv.org/abs/2410.13569" target="_blank">Paper</a> | 💻 <a href="https://github.com/eliahuhorwitz/ProbeX" target="_blank">GitHub</a> | 🤗 <a href="https://huggingface.co/ProbeX" target="_blank">Dataset</a>
</p>

![ProbeX](https://raw.githubusercontent.com/eliahuhorwitz/ProbeX/main/imgs/poster.png)

## Model Details

| Attribute | Value |
|---|---|
| **Subset** | ResNet |
| **Split** | train |
| **Base Model** | `microsoft/resnet-101` |
| **Dataset** | CIFAR100 (50 classes) |

## Training Hyperparameters

| Parameter | Value |
|---|---|
| Learning Rate | 0.0001 |
| LR Scheduler | constant |
| Epochs | 7 |
| Max Train Steps | 2331 |
| Batch Size | 64 |
| Weight Decay | 0.009 |
| Seed | 433 |
| Random Crop | True |
| Random Flip | False |

## Performance

| Metric | Value |
|---|---|
| Train Accuracy | 0.9627 |
| Val Accuracy | 0.8595 |
| Test Accuracy | 0.8570 |

## Training Categories

The model was fine-tuned on the following 50 CIFAR100 classes:

`baby`, `maple_tree`, `can`, `otter`, `motorcycle`, `skyscraper`, `mushroom`, `leopard`, `crab`, `bowl`, `aquarium_fish`, `seal`, `bus`, `road`, `forest`, `worm`, `shrew`, `butterfly`, `oak_tree`, `clock`, `bee`, `chair`, `beaver`, `boy`, `orchid`, `snail`, `hamster`, `fox`, `bridge`, `pine_tree`, `cloud`, `camel`, `woman`, `snake`, `squirrel`, `girl`, `tulip`, `poppy`, `possum`, `television`, `dolphin`, `wolf`, `lion`, `lobster`, `bicycle`, `tractor`, `rabbit`, `pickup_truck`, `couch`, `flatfish`