Table of Contents

OpenAI Whisper

About Whisper

Problems / Forums

Setup

1. Upgrade to Jetpack 5.1

This means basically a completely new installation / initial setup.
Follow Getting Started with Jetson AGX Orin Developer Kit to flash the Orin new.
Use the NVIDIA SDK Manager to setup the Orin via USB from scratch.

2. Install pip and virtualenv for Python3

sudo apt install python3-pip
sudo apt install virtualenv

# as user upgrade pip
pip install --upgrade pip

3. Install PyTorch for Jetson

:!: Follow Install-PyTorch-Jetson-Platform.pdf

Alternative howto: https://docs.nvidia.com/deeplearning/frameworks/install-pytorch-jetson-platform/index.html

Prepare

sudo apt-get -y update

sudo apt-get -y install \
autoconf bc build-essential g++-8 gcc-8 clang-8 lld-8 gettext-base gfortran-8 \
iputils-ping libbz2-dev libc++-dev libcgal-dev libffi-dev libfreetype6-dev libhdf5-dev \
libjpeg-dev liblzma-dev libncurses5-dev libncursesw5-dev libpng-dev libreadline-dev libssl-dev \
libsqlite3-dev libxml2-dev libxslt-dev locales moreutils openssl python-openssl rsync scons \
python3-pip virtualenv libopenblas-dev

Check for the latest version of PyTorch for Jetpack:
https://developer.download.nvidia.com/compute/redist/jp/v51/pytorch/

Save the result in an environment variable:

export TORCH_INSTALL=https://developer.download.nvidia.com/compute/redist/jp/v51/pytorch/torch-1.14.0a0+44dac51c.nv23.02-cp38-cp38-linux_aarch64.whl

Install PyTorch in the environment nlp

# create and activate the environment nlp
virtualenv -p python3 nlp
source nlp/bin/activate

# pip install aiohttp numpy=='1.19.4' scipy=='1.5.3' 
pip install aiohttp numpy scipy 
export "LD_LIBRARY_PATH=/usr/lib/llvm-8/lib:$LD_LIBRARY_PATH"
pip install --upgrade protobuf; 
pip install --no-cache $TORCH_INSTALL

Install Jupyterlab and Pandas

pip install pandas matplotlib jupyterlab

Install Whisper by OpenAI

pip install -U openai-whisper

In case you want to delete the installation:

deactivate
rm -Rf nlp

Test OpenAI Whisper

source nlp/bin/activate

pip install ipywidgets 

pip install torchaudio

Not working: Upgrade to Jetpack 5.1, r35.2


Bummer! This upgrade did not work! It was corrupt!
The Orin did not boot anymore!
Workaround: Use the NVIDIA SDK Manager to setup the Orin via USB from scratch!

Install from NVIDIA Jetson Repository: https://repo.download.nvidia.com/jetson/

sudo bash -c 'echo "deb https://repo.download.nvidia.com/jetson/common r35.2 main" >> /etc/apt/sources.list.d/nvidia-l4t-apt-source.list'
sudo bash -c 'echo "deb https://repo.download.nvidia.com/jetson/t234 r35.2 main" >> /etc/apt/sources.list.d/nvidia-l4t-apt-source.list'
sudo apt update
sudo apt dist-upgrade
sudo apt install --fix-broken -o Dpkg::Options::="--force-overwrite"
root@agx-orin:/etc/apt/sources.list.d# apt list --upgradable
Listing... Done
jetson-gpio-common/stable 2.0.210ubuntu1 arm64 [upgradable from: 2.0.180ubuntu1]
libopencv-dev/stable 4.5.4-8-g3e4c170df4 arm64 [upgradable from: 4.2.0+dfsg-5]
nvidia-l4t-3d-core/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-apt-source/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-bootloader/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-camera/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-configs/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-core/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-cuda/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-display-kernel/stable 5.10.104-tegra-35.2.1-20230124153320 arm64 [upgradable from: 5.10.65-tegra-34.0.1-20220315011106]
nvidia-l4t-firmware/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-gputools/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-graphics-demos/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-gstreamer/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-init/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-initrd/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-jetson-io/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-kernel-dtbs/stable 5.10.104-tegra-35.2.1-20230124153320 arm64 [upgradable from: 5.10.65-tegra-34.0.1-20220315011106]
nvidia-l4t-kernel-headers/stable 5.10.104-tegra-35.2.1-20230124153320 arm64 [upgradable from: 5.10.65-tegra-34.0.1-20220315011106]
nvidia-l4t-kernel/stable 5.10.104-tegra-35.2.1-20230124153320 arm64 [upgradable from: 5.10.65-tegra-34.0.1-20220315011106]
nvidia-l4t-libvulkan/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-multimedia-utils/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-multimedia/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-nvfancontrol/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-nvpmodel-gui-tools/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-nvpmodel/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-nvsci/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-oem-config/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-optee/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-pva/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-tools/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-wayland/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-weston/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-x11/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
nvidia-l4t-xusb-firmware/stable 35.2.1-20230124153320 arm64 [upgradable from: 34.0.1-20220315011106]
python-jetson-gpio/stable 2.0.210ubuntu1 arm64 [upgradable from: 2.0.180ubuntu1]
python3-jetson-gpio/stable 2.0.210ubuntu1 arm64 [upgradable from: 2.0.180ubuntu1]
root@agx-orin:/etc/apt/sources.list.d# 
Setting up nvidia-l4t-bootloader (35.2.1-20230124153320) ...
3701--0000--1--jetson-agx-orin-devkit-
Info. Active boot storage: mmcblk0
Info. Legacy mode: true
INFO. Dump nv_boot_control.conf:
TNSPEC 3701-500-0000-M.0-1-1-jetson-agx-orin-devkit-mmcblk0p1
COMPATIBLE_SPEC 3701--0000--1--jetson-agx-orin-devkit-
TEGRA_LEGACY_UPDATE true
TEGRA_BOOT_STORAGE mmcblk0
TEGRA_EMMC_ONLY false
TEGRA_CHIPID 0x23
TEGRA_OTA_BOOT_DEVICE /dev/mtdblock0
TEGRA_OTA_GPT_DEVICE /dev/mtdblock0
Starting bootloader post-install procedure.
Starting legacy update procedure.





Update bl_only_payload completed.
Updating esp partition...
done.
Reboot the target system for changes to take effect.
Updating extlinux.conf...
Root device is set in the extlinux.conf
Setting up nvidia-l4t-nvsci (35.2.1-20230124153320) ...
Setting up nvidia-l4t-gputools (35.2.1-20230124153320) ...
Setting up nvidia-l4t-wayland (35.2.1-20230124153320) ...
Setting up jetson-gpio-common (2.0.210ubuntu1) ...
Setting up nvidia-l4t-xusb-firmware (35.2.1-20230124153320) ...
Setting up nvidia-l4t-cuda (35.2.1-20230124153320) ...
Setting up nvidia-l4t-optee (35.2.1-20230124153320) ...
Setting up nvidia-l4t-kernel (5.10.104-tegra-35.2.1-20230124153320) ...
Using the existing boot entry 'primary'
3701--0000--1--jetson-agx-orin-devkit-
Info. Active boot storage: mmcblk0
Info. Legacy mode: true
Starting kernel post-install procedure.
Starting legacy update procedure.
ERROR. Procedure for A_kernel-dtb update FAILED.
Cannot install package. Exiting...
dpkg: error processing package nvidia-l4t-kernel (--configure):
 installed nvidia-l4t-kernel package post-installation script subprocess returned error exit status 1
Setting up nvidia-l4t-nvpmodel-gui-tools (35.2.1-20230124153320) ...
Setting up python-jetson-gpio (2.0.210ubuntu1) ...
Setting up nvidia-l4t-nvfancontrol (35.2.1-20230124153320) ...
Installing new version of config file /etc/nvpower/nvfancontrol/nvfancontrol_e3900.conf ...
Installing new version of config file /etc/nvpower/nvfancontrol/nvfancontrol_p2888.conf ...
Installing new version of config file /etc/nvpower/nvfancontrol/nvfancontrol_p3668.conf ...
Installing new version of config file /etc/nvpower/nvfancontrol/nvfancontrol_p3701_0000.conf ...
Installing new version of config file /etc/systemd/system/nvfancontrol.service ...
dpkg: dependency problems prevent configuration of nvidia-l4t-kernel-headers:
 nvidia-l4t-kernel-headers depends on nvidia-l4t-kernel (= 5.10.104-tegra-35.2.1-20230124153320); however:
  Package nvidia-l4t-kernel is not configured yet.

dpkg: error processing package nvidia-l4t-kernel-headers (--configure):
 dependency problems - leaving unconfigured
dpkg: dependency problems prevent configuration of nvidia-l4t-jetson-io:
 nvidia-l4t-jetson-io depends on nvidia-l4t-kernel (>> 5.10.104-tegra-35.2-0); however:
  Package nvidia-l4t-kernel is not configured yet.
 nvidia-l4t-jetson-io depends on nvidia-l4t-kernel (<< 5.10.104-tegra-35.3-0); however:
  Package nvidia-l4t-kernel is not configured yet.

No apport report written because the error message indicates its a followup error from a previous failure.
                                                                                                          No apport report written because the error message indicates its a followup error from a previous
 failure.
         dpkg: error processing package nvidia-l4t-jetson-io (--configure):
 dependency problems - leaving unconfigured
Setting up nvidia-l4t-x11 (35.2.1-20230124153320) ...
Installing new version of config file /etc/X11/xorg.conf.t194_ref ...
Setting up nvidia-l4t-3d-core (35.2.1-20230124153320) ...
Setting up python3-jetson-gpio (2.0.210ubuntu1) ...
dpkg: dependency problems prevent configuration of nvidia-l4t-display-kernel:
 nvidia-l4t-display-kernel depends on nvidia-l4t-kernel (= 5.10.104-tegra-35.2.1-20230124153320); however:
  Package nvidia-l4t-kernel is not configured yet.

dpkg: error processing package nvidia-l4t-display-kernel (--configure):
 dependency problems - leaving unconfigured
Setting up nvidia-l4t-initrd (35.2.1-20230124153320) ...
No apport report written because MaxReports is reached already
                                                              Setting up nvidia-l4t-multimedia (35.2.1-20230124153320) ...
dpkg: dependency problems prevent configuration of nvidia-l4t-kernel-dtbs:
 nvidia-l4t-kernel-dtbs depends on nvidia-l4t-kernel (= 5.10.104-tegra-35.2.1-20230124153320); however:
  Package nvidia-l4t-kernel is not configured yet.

dpkg: error processing package nvidia-l4t-kernel-dtbs (--configure):
 dependency problems - leaving unconfigured
Setting up nvidia-l4t-pva (35.2.1-20230124153320) ...
No apport report written because MaxReports is reached already
                                                              Setting up nvidia-l4t-gbm (35.2.1-20230124153320) ...
Setting up nvidia-l4t-camera (35.2.1-20230124153320) ...
Setting up nvidia-l4t-gstreamer (35.2.1-20230124153320) ...
Setting up nvidia-l4t-weston (35.2.1-20230124153320) ...
Setting up nvidia-l4t-graphics-demos (35.2.1-20230124153320) ...
Processing triggers for libc-bin (2.31-0ubuntu9.9) ...
Errors were encountered while processing:
 nvidia-l4t-kernel
 nvidia-l4t-kernel-headers
 nvidia-l4t-jetson-io
 nvidia-l4t-display-kernel
 nvidia-l4t-kernel-dtbs
E: Sub-process /usr/bin/dpkg returned an error code (1)