Model compression via Pruning and Quantization on NVIDIA Jetson Nano

Paper

This repository contains source code for model compression techniques on the NVIDIA Jetson Nano, focusing on pruning and quantization.

Dataset & Model

Given the resource constraints of the Jetson Nano, the following dataset and model were selected:

• Dataset: FashionMNIST
• Model : LeNet-5

Model Compression techniques

• Pruning
• Quantization

Code

Install dependencies

python -m pip install -r requirements.txt

Experimentation

All experimental code can be found in experiment.ipynb. This notebook contains step-by-step implementations of pruning and quantization techniques on the selected model and dataset.