Prerequisites

• Python 3.8 or above installed
• pip for installing Python packages

---

Setup Instructions

1. Clone the Repository

git clone <your-repository-url>
cd <your-repository-folder>

2. Install Dependencies

pip install -r requirements.txt

Deliverables and Usage

A. Model Preprocessing and Inference Code

• pytorch_model.py: Contains the PyTorch classifier definition and image preprocessing methods.
• onnx_model.py: Loads the ONNX model and runs inference.

These modules are used internally and for local testing.

---

B. test.py — Local Model Testing Script

Use this to test your model locally before deployment.

python test.py <path_to_image>

It prints the predicted class ID for the given image.

C. test_server.py — Testing the Deployed Model

This script tests your deployed model on Cerebrium by sending images and checking predictions.

Usage

To predict an image class ID:

python test_server.py <path_to_image>

To run platform monitoring tests:

python test_server.py --run-tests

Sources

• Pytorch Documentation to Convert to ONNX model - [Link to Documentation]
• Pytorch Documentation - [Link to Documentation]
• ONNX Load and Run Documentation - [Link to Documentation]
• Cerebrium Documentation - [Link to Documentation]

Contact

ahmedirtija8@gmail.com