PyTorch for Jetson Nano - Shell/Bash

Introduction

PyTorch is an open source machine learning library that is widely used in research and industry. It provides a simple and easy-to-use interface for building and training neural networks, making it an ideal tool for both beginners and experts in the field of deep learning.

Jetson Nano is a small, powerful computer designed for embedded applications and AI on the edge. It has a quad-core ARM processor and a powerful NVIDIA GPU that makes it perfect for running deep learning models using PyTorch.

In this tutorial, we'll take a look at how to install and use PyTorch on the Jetson Nano using the Shell/Bash command line.

Prerequisites

Before we start, you'll need the following:

• A Jetson Nano developer kit
• A microSD card with JetPack installed
• An Internet connection

Installing PyTorch

To install PyTorch, we will use pip, the default package installer for Python. Open a terminal window on your Jetson Nano and run the following commands:

sudo apt-get install python3-pip libopenblas-dev libblas-dev m4 cmake cython python3-dev python3-yaml python3-setuptools
sudo pip3 install numpy torch-1.6.0-cp36-cp36m-linux_aarch64.whl torchaudio-0.6.0-cp36-cp36m-linux_aarch64.whl torchvision-0.7.0-cp36-cp36m-linux_aarch64.whl

This will install the necessary dependencies and then install PyTorch and its associated libraries.

Testing Installation

Once PyTorch is installed, we can test to make sure everything is working correctly. Open a Python shell by typing "python3" in the terminal and run the following commands:

import torch
print(torch.__version__)
print(torch.cuda.is_available())

This should output the version number of PyTorch and "True" if a CUDA-compatible GPU is available.

Conclusion

In this tutorial, we looked at how to install and use PyTorch on the Jetson Nano using the Shell/Bash command line. With PyTorch, we can now easily build and train neural networks on the Jetson Nano, making it a powerful tool for AI on the edge.