NVIDIA now has an official release for TensorFlow on the NVIDIA Jetson TX2 Development Kit!
This makes installing TensorFlow on the Jetson much less challenging. Here’s the shortcut version:
For Python 2.7
$ pip install –extra-index-url=https://developer.download.nvidia.com/compute/redist/jp33 tensorflow-gpu
For Python 3.5
pip3 install –extra-index-url=https://developer.download.nvidia.com/compute/redist/jp33 tensorflow-gpu
Here is the original announcement and the full installation document.
Here are some other useful links
Thanks, Jim, I was having a hard time trying to install TF in my TX2 using the previous method. This one works like a charm.
I’m glad you got it to work. Thanks for reading!
Would it be possible to get simple installs like thing for TX1
This should work for you.
Invalid requirement: ‘–extra-index-url=https://developer.download.nvidia.com/compute/redist/jp33’
It looks like a path. File ‘–extra-index-url=https://developer.download.nvidia.com/compute/redist/jp33’ does not exist.
I can’t use it.Why?
Please ask this question here: https://devtalk.nvidia.com/default/topic/1038957/jetson-tx2/tensorflow-for-jetson-tx2-/post/5278617/#5278617
Above is a NVIDIA project, please ask them about issues.
Thanks for making everyone Life so much better.
I also wanted to let everyone know that here is another Lecture that you can take online.
Module Video Material
Main video lecture (complete YouTube Playlist):
Part 1.1: Course Overview
Part 1.2: Machine Learning Background for Deep Learning, Keras and Tensorflow
Part 1.3: Python Anaconda for Deep Learning, Keras and Tensorflow
How to Submit a Module Assignment
Watch one (or more) of these depending on how you want to setup your Python TensorFlow environment:
Installing TensorFlow, Keras, and Python in Windows
Installing TensorFlow, Keras, and Python in Mac
Installing/Using IBM Cognitive Class Labs with TensorFlow/Keras
Docker Image – A docker image that I created specifically for this class. Always tested with all class assignments and notes.
Thanks for the kind words, links, and thanks for reading!
Hello,Jim,I just install Jetpack4.1.1 on my nvidia xavier and then wanna install the tensorflow on the board.Thus I follow the official docunment:
Unfortunately comfront some errors.
Firstly,I install Nvidia SDK Manager:
$ sudo apt install ./sdkmanager_0.9.11-3405_amd64.deb
Some packages could not be installed. This may mean that you have
requested an impossible situation or if you are using the unstable
distribution that some required packages have not yet been created
or been moved out of Incoming.
The following information may help to resolve the situation:
The following packages have unmet dependencies:
sdkmanager:amd64 : Depends: libgconf-2-4:amd64 but it is not installable
Depends: libcanberra-gtk-module:amd64 but it is not installable
E: Unable to correct problems, you have held broken packages.
$ sudo apt install libgconf-2-4
libgconf-2-4 is already the newest version (3.2.6-4ubuntu1).
The following packages were automatically installed and are no longer required:
libdbusmenu-gtk4 libdbusmenu-qt5-2 libgsettings-qt1 liblockfile-bin
liblockfile1 libqt5sql5 libqt5sql5-sqlite lockfile-progs x11proto-dri2-dev
Use ‘sudo apt autoremove’ to remove them.
The dependencies are exist,but I still couldn’t install SDK. Why?
Appreciate to receive your answers,thanks.
NVIDIA now supports TensorFlow officially: https://devtalk.nvidia.com/default/topic/1042125/jetson-agx-xavier/official-tensorflow-for-jetson-agx-xavier/
So that we can better share this information with the community, could you please ask this question on the official NVIDIA Jetson forum where a large number of developers and NVIDIA engineers share their experience. That way everyone can benefit from the answer.
The Jetson AGX Xavier forum is here: https://devtalk.nvidia.com/default/board/326/jetson-agx-xavier/
Thanks for reading!
Thx for the great article!
Unluckily the NVIDIA l4t docker image (as supported by JetPack 4.2.1) or pip packages do not contain TensorFlow *Serving* which is quite interesting for inferencing on the edge.
Have a look at https://github.com/helmuthva/jetson/tree/master/workflow/deploy/tensorflow-serving-base/src for a Dockerfile and .bazelrc to build the latest TensorFlow Serving (inc. TensorFlow Core) from master for NVIDIA Jetson devices.
See https://github.com/helmuthva/jetson for the bigger picture – a multi-arch Kubernetes cluster with edge devices for inferencing.
I have been watching your Github repository for a while now, very nice work! It seems the whole container area isn’t very well explained/explored on the Jetson currently. Hopefully your work helps clarify some of these issues. Also, you should put your project in the ‘Projects’ area in the ‘Jetson Projects’ area of the NVIDIA forums. That way it will get more exposure from people who are trying to spread the word. Thanks for reading!