Jetson Nano Setup Guide

Introduction

Welcome to the Jetson Nano Setup Guide! This guide will walk you through the process of setting up your Jetson Nano developer kit, from downloading the system image to configuring advanced usage scenarios.

Download Jetson Nano Image

Before you begin, you’ll need to download the system image for your Jetson Nano model:

• Jetson Nano 2GB
• Jetson Nano 4GB

Setting up the MicroSD with Ubuntu using balenaEther

1. Go to this website
2. Find and download Jetson Nano Developer Kit SD Card Image
3. Download and install balenaEtcher
4. Open balenaEthcer
   1. Select: Flash from file
   2. Select: jetson-nano-jp461-sd-card-image.zip
   3. Select the MicroSD card
   4. Flash!
5. Place the MicroSD back into the jetson nano and start.

Fresh Install Setup Guide for Jetson Nano

Once your Jetson Nano board is up and running with Ubuntu Desktop, let’s kickstart the setup process by opening the terminal and following these steps:

Step 1: Update System

First, let’s ensure your system is up to date:

sudo apt update

Next, upgrade your Jetson Nano:

sudo apt upgrade -y

During the upgrade process, you may encounter prompts for configuration files.

Select Y to install the package maintainer’s version.

Configuration file '/etc/ld.so.conf.d/nvidia-tegra.conf'
 ==> Modified (by you or by a script) since installation.
 ==> Package distributor has shipped an updated version.
   What would you like to do about it ?  Your options are:
    Y or I  : install the package maintainer's version
    N or O  : keep your currently-installed version
      D     : show the differences between the versions
      Z     : start a shell to examine the situation
 The default action is to keep your current version.
*** nvidia-tegra.conf (Y/I/N/O/D/Z) [default=N] ? Y

Again select Y to install the package maintainer’s version.

Configuration file '/etc/systemd/nv-oem-config-post.sh'
 ==> Deleted (by you or by a script) since installation.
 ==> Package distributor has shipped an updated version.
   What would you like to do about it ?  Your options are:
    Y or I  : install the package maintainer's version
    N or O  : keep your currently-installed version
      D     : show the differences between the versions
      Z     : start a shell to examine the situation
 The default action is to keep your current version.
*** nv-oem-config-post.sh (Y/I/N/O/D/Z) [default=N] ? Y

Additionally, if prompted to restart Docker, select YES.

Now, let’s perform a distribution upgrade:

sudo apt dist-upgrade -y

Step 2: Clean Up

Once the upgrade process is complete, let’s tidy up by removing old packages:

sudo apt autoremove -y

And finally, let’s clean up the cache:

sudo apt clean

Step 3: Reboot

After these maintenance tasks, it’s recommended to reboot your Jetson Nano for changes to take effect:

sudo reboot now

Install Useful Tools

Here are some essential tools that are handy for almost every project:

sudo apt install git nano curl wget -y

Uninstall LibreOffice

If you no longer need LibreOffice and want to reclaim some disk space, follow these steps to remove it:

sudo apt autoremove libreoffice* -y

This command will uninstall all LibreOffice packages from your system.

After removing LibreOffice, let’s clean up the residual files:

sudo apt clean

This command will clean the package cache, freeing up additional disk space.

Your system is now free of LibreOffice and optimized for your needs.

Installing pip and pip3

To install both pip and pip3, which are package managers for Python 2 and Python 3 respectively, run the following command:

sudo apt install python-pip python3-pip -y

This command will install pip for Python 2 and pip3 for Python 3 on your system.

You’re all set with pip and pip3 installed and ready to manage Python packages!

Installing Jetson Stats

To install Jetson Stats, a utility for monitoring and controlling NVIDIA Jetson devices, follow these steps:

sudo pip3 install -U jetson-stats

This command will install Jetson Stats and ensure that you have the latest version.

After installation, reboot your Jetson Nano to enable the jtop command:

sudo reboot now

Once your device has rebooted, reopen the terminal and type the following command to launch Jetson Stats:

jtop

This will open the Jetson Stats interface, allowing you to monitor various aspects of your Jetson Nano’s performance.

You’re now ready to utilize Jetson Stats for optimizing your Jetson Nano’s performance!

Configuring Jetson Fan to Start at Boot

To ensure your Jetson Nano’s fans start automatically at boot, follow these steps:

1. Open and edit the rc.local file using the Nano text editor:

sudo nano /etc/rc.local

Paste the following lines into the file:

#!/bin/bash
sleep 10
sudo /usr/bin/jetson_clocks
sudo sh -c 'echo 255 > /sys/devices/pwm-fan/target_pwm'
exit 0

2. Next, create and edit the rc-local.service file:

sudo nano /etc/systemd/system/rc-local.service

Insert the following content:

[Unit]
Description=/etc/rc.local Compatibility
ConditionPathExists=/etc/rc.local

[Service]
Type=forking
ExecStart=/etc/rc.local start
TimeoutSec=0
StandardOutput=tty
RemainAfterExit=yes
SysVStartPriority=99

[Install]
WantedBy=multi-user.target

3. Ensure the rc.local file has execute permissions:

sudo chmod +x /etc/rc.local

4. Now, enable the rc-local service to run on system boot:

sudo systemctl enable rc-local

5. Start the rc-local service:

sudo systemctl start rc-local.service

Your Jetson Nano’s fans will now start automatically at boot, ensuring optimal cooling performance.

Increase Swap Space

To increase the swap space on your Jetson Nano, follow these steps:

1. Clone the resizeSwapMemory repository:

git clone https://github.com/JetsonHacksNano/resizeSwapMemory

2. Navigate to the cloned repository:

cd resizeSwapMemory

3. Set the entire swap memory size to 8GB:

./setSwapMemorySize.sh -g 8

This will ensure sufficient memory for the OpenCV installation.

OpenCV Installation Guide

Before installing OpenCV on your Jetson Nano, ensure that your system has sufficient memory by following the steps above to increase swap space if needed.

Next, download the OpenCV installation script:

wget https://github.com/Qengineering/Install-OpenCV-Jetson-Nano/raw/main/OpenCV-4-8-0.sh

Set the appropriate permissions for the script:

sudo chmod 755 ./OpenCV-4-8-0.sh

Run the installation script:

./OpenCV-4-8-0.sh

Once the installation is complete, you can remove the installation script:

rm OpenCV-4-8-0.sh

Finally, you can remove the dphys-swapfile:

sudo /etc/init.d/dphys-swapfile stop
sudo apt-get remove --purge dphys-swapfile

As a tip to save additional space, you can remove the OpenCV and OpenCV_contrib directories:

sudo rm -rf ~/opencv
sudo rm -rf ~/opencv_contrib

Congratulations! You’ve successfully installed OpenCV on your Jetson Nano.

Install Python 3.12 on Jetson Nano From Source

Python 3.12 brings new features, improvements, and optimizations to the language, making it desirable for developers who want to leverage the latest capabilities.

To install Python 3.12 on your Jetson Nano from source, follow these steps:

1. Update your system’s package list:

sudo apt update

2. Install the necessary dependencies for building Python:

sudo apt install wget build-essential zlib1g-dev libncurses5-dev libgdbm-dev libnss3-dev libssl-dev libreadline-dev libffi-dev -y

3. Create a directory for the Python source code and navigate to it:

mkdir ./python && cd ./python

4. Download the Python source code. Replace 3.12.0 with the desired Python version:

wget https://www.python.org/ftp/python/3.12.0/Python-3.12.0.tgz

5. Extract the downloaded archive:

tar -xvf Python-3.12.0.tgz

6. Navigate to the extracted directory:

cd Python-3.12.0

7. Configure the build with optimizations enabled:

./configure --enable-optimizations

8. Compile and install Python:

sudo make install

9. After completing the installation, reboot your Jetson Nano:

sudo reboot now

After rebooting, Python 3 will be installed on your Jetson Nano from the source code. You can verify the installation by running python3 --version.

Install CrewAI on Jetson Nano

CrewAI is a powerful AI platform designed to assist with a variety of tasks. To install CrewAI on your Jetson Nano, follow these steps:

To install the main CrewAI package, which includes core functionalities, run the following command:

pip3 install crewai

If you also want to install the tools package, which includes a series of helpful tools for your agents, you can use the following command:

pip3 install 'crewai[tools]'

This command will install the main CrewAI package along with additional tools to enhance your CrewAI experience.

Once installed, you can start using CrewAI to develop and deploy AI solutions on your Jetson Nano.

Make test script for crewAI (Testing)

mkdir test_ai && cd test_ai

nano .env

and then add this

ANTHROPIC_API_KEY=your_anthropic_api_key

Now save and close nano

now let’s create our main.py

nano main.py

And then add this code below

import os
from dotenv import load_dotenv
from crewai import Agent, Task, Crew
from langchain.llms import Anthropic
from textwrap import dedent

# Load the API key from the .env file
load_dotenv()
ANTHROPIC_API_KEY = os.getenv("ANTHROPIC_API_KEY")

# Define a simple agent
class SimpleAgent:
    def __init__(self):
        self.Claude = Anthropic(anthropic_api_key=ANTHROPIC_API_KEY, temperature=0.7)

    def create_agent(self):
        return Agent(
            role="Simple Test Agent",
            backstory=dedent("""You are a helpful assistant."""),
            goal=dedent("""Your goal is to assist the user with their query."""),
            allow_delegation=False,
            verbose=True,
            llm=self.Claude,
        )

# Define a simple task
class SimpleTask:
    def __init__(self, user_query):
        self.user_query = user_query

    def create_task(self, agent):
        return Task(
            description=dedent(
                f"""
            Please assist with the following query:
            
            {self.user_query}
        """
            ),
            agent=agent,
        )

# Main function
if __name__ == "__main__":
    print("## Welcome to the Crew AI Test")
    print("-------------------------------")

    user_query = input(dedent("""Enter your query: """))

    simple_agent = SimpleAgent().create_agent()
    simple_task = SimpleTask(user_query).create_task(simple_agent)

    crew = Crew(agents=[simple_agent], tasks=[simple_task], verbose=True)
    result = crew.kickoff()

    print("\n\n########################")
    print("## Here is the result:")
    print("########################\n")
    print(result)