install_steps.md

Basic Installation

1. Etch the Ubuntu Image using the instructions here
2. Configure the installation (requires a keyboard/mouse) after the reboot.
3. sudo apt update && sudo apt upgrade
4. sudo apt install python3-pip python-pip
5. Reboot Jetson Nano
6. cd ~
7. wget https://github.com/conda-forge/miniforge/releases/latest/download/Miniforge3-Linux-aarch64.sh .
8. chmod a+x Miniforge3-Linux-aarch64.sh
9. ./Miniforge3-Linux-aarch64.sh (Do not run as root!)
10. Log out and log back on
11. conda config --set auto_activate_base false
12. sudo apt install python3-h5py libhdf5-serial-dev hdf5-tools libpng-dev libfreetype6-dev
13. conda create -n jupyter python=3.6
14. conda activate jupyter 1. conda install matplotlib pandas numpy pillow scipy tqdm scikit-image scikit-learn seaborn cython h5py jupyter ipywidgets -c conda-forge

Optional Installation

• Pytorch

  • https://forums.developer.nvidia.com/t/pytorch-for-jetson-version-1-10-now-available/72048
  • wget https://nvidia.box.com/shared/static/fjtbno0vpo676a25cgvuqc1wty0fkkg6.whl -O torch-1.10.0-cp36-cp36m-linux_aarch64.whl
  • sudo apt-get install libopenblas-base libopenmpi-dev
  • pip install torch-1.10.0-cp36-cp36m-linux_aarch64.whl

• torchvision:

  • https://forums.developer.nvidia.com/t/pytorch-for-jetson-version-1-10-now-available/72048
  • sudo apt-get install libjpeg-dev zlib1g-dev libpython3-dev libavcodec-dev libavformat-dev libswscale-dev
  • git clone --branch v0.11.1 https://github.com/pytorch/vision torchvision
  • cd torchvision
  • export BUILD_VERSION=0.11.1
  • python setup.py install --user

• TensorRT:

  • sudo apt-get install tensorrt
  • export PYTHONPATH=/usr/lib/python3.6/dist-packages:$PYTHONPATH
  • pip install tensorrt

• TensorFlow:

  • https://docs.nvidia.com/deeplearning/frameworks/install-tf-jetson-platform/index.html#install
  • pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v46 tensorflow

General Tips and Tricks

• Jetson Nano CLI Only:

  • If you're only using the Jetson Nano for non-gui purposes, like as a remote inference/training server. This reduces memory pressure of running your OS, and instead focuses on leaving enough memory for your training/inference operations.
  • jetson_config -p desktop
  • Select [B1] to boot directly to non-gui terminal and require login.
  • NOTE: Only use [B2] if you understand what the consequences of this option.
  • This will trigger a Jetson Nano reboot. Click yes.

• Other Useful Utilities:

  • htop: Improved top command, provides instantaneous stats about your machine/board.
    • https://htop.dev/
    • sudo apt install htop
  • tmux: Terminal Multiplexer. Very useful if you want to start running a command on one terminal instance and check on it periodically on another.
    • https://www.hamvocke.com/blog/a-quick-and-easy-guide-to-tmux/
    • sudo apt install tmux
  • Locate: Find files easily and instantly.
    • https://linuxize.com/post/locate-command-in-linux/
    • sudo apt install mlocate
  • Jetson Stats : More detailed information about your Jetson board
    • https://github.com/rbonghi/jetson_stats
    • sudo -H pip3 install -U jetson-stats
    • sudo systemctl restart jetson_stats.service
    • jtop
    • Reboot your Jetson Nano.

• Disk Imager:

  • Creating a working image for your jetson nano board.
  • windows: https://win32diskimager.org/#download

• Cleanup:

  • sudo apt autoremove