/jetcard

An SD card image for web programming AI projects with NVIDIA Jetson Nano

Primary LanguagePythonMIT LicenseMIT

JetCard

JetCard is a system configuration that makes it easy to get started with AI. It comes pre-loaded with

  • A Jupyter Lab server that starts on boot for easy web programming

  • A script to display the Jetson's IP address (and other stats)

  • The popular deep learning frameworks PyTorch and TensorFlow

After configuring your system using JetCard, you can get started prototyping AI projects from your web browser in Python.

If you find an issue, please let us know!

Setup

Follow the steps below to download JetCard directly or create it from scratch.

Option 1 - Download JetCard directly

  1. Download the JetCard image jetcard_v0p0p0.img onto a Windows, Linux or Mac desktop machine

    You can check it against this md5sum

  2. Insert a 32GB+ SD card into the desktop machine

  3. Using Etcher select jetcard_v0p0p0.img and flash it onto the SD card

  4. Remove the SD card from the desktop machine

You may now insert the SD card into the Jetson Nano, power on, and enjoy the pre-configured system!

Please note, the password for the pre-built SD card is jetson

Option 2 - Create JetCard from scratch

  1. Flash Jetson Nano following the Getting Started Guide

    For Jetson TX2 / Xavier, use the JetPack SDK manager

  2. On the Jetson, run the JetCard installation script

    git clone https://github.com/NVIDIA-AI-IOT/jetcard
    cd jetcard
    ./install.sh <password>

Once the install.sh script finishes, your system should be configured identically to the SD card image mentioned above.

Usage

Connecting

Pick an option below and follow the instructions to begin web programming Jetson from a desktop computer using Jupyter Lab.

Option 1 - Ethernet / WiFi

  1. Power on the Jetson platform configured using JetCard

  2. Connect the Jetson to the same network as your desktop computer via Ethernet or WiFi

    If you want to connect your Jetson to WiFi, but don't have a monitor and keyboard, you can connect via device mode (see below), open a terminal, and then use the nmcli tool to connect to a WiFi network. Find more details below.

  3. Determine the IP address jetson_ip_address

    If you have the PiOLED display attached, it will display on that screen. Otherwise, you will need to connect a monitor, open a terminal, and read the IP using ifconfig.

  4. Connect to the Jetson platform from a desktop computer by navigating to http://<jetson_ip_address>:8888

  5. Sign in using the default password jetson

Option 2 - USB device mode

If you do not occupy the Jetson Nano's micro USB port for power, you can use it to connect directly from a desktop PC! The USB device mode IP address is 192.168.55.1

  1. Power on the Jetson platform configured using JetCard

  2. Connect the Jetson platform to the desktop computer via micro USB

  3. On the desktop computer, navigate to http://192.168.55.1:8888 from a web browser

  4. Sign in using the default password jetson

Extras

Connect to WiFi from terminal

To connect your Jetson to a WiFi network from a terminal, follow these steps

  1. Re-scan available WiFi networks

    nmcli device wifi rescan
  2. List available WiFi networks, and find the ssid_name of your network.

    nmcli device wifi list
  3. Connect to a selected WiFi network

    nmcli device wifi connect <ssid_name> password <password>

Create SD card snapshot

If you've applied modifications to the base SD card image that you want to re-use, do the following to create a compressed SD card image

  1. Remove the SD card from your Jetson Nano

  2. Insert the SD card into a Linux host computer

  3. Determine where the SD card is located using sudo fdisk -l. We'll assume this is at /dev/sdb

  4. Copy the contents of the SD card to a file named jetcard_image.img

    sudo dd bs=4M if=/dev/sdb of=jetcard_image.img status=progress
  5. Compress the SD card image using zip

    zip jetcard_image.zip jetcard_image.img

See also

  • JetBot - An educational AI robot based on NVIDIA Jetson Nano

  • JetRacer - An educational AI racecar using NVIDIA Jetson Nano

  • JetCam - An easy to use Python camera interface for NVIDIA Jetson

  • torch2trt - An easy to use PyTorch to TensorRT converter