Is there a way to do Getting Started with AI on Jetson Nano with a Xavier

I got a Xavier not a Nano. Is there a way to follow the Getting Started with AI on Jetson Nano course with the Xavier?

Hi,

The procedure is roughly the same.
You can just follow the course.

Thanks.

But with the nano, we get to use a special version of the micro-SD, don’t we? ( I recall that it uses jupyter note etc). The problem would be how to do the setup for the Xavier?

Hi @aliaga, the updated version of the DLI course uses containers, not custom SD card image. So you should be able to run the container on Xavier too.

Oh really?! great! I ll make sure to review that! Great to hear that! And thanks a lot!

can we re-open this because im seeing that its not true that you can run the DLI course , using containers , on a Xavier.

Heres a question where i documented my troubles.

here is the docker run command i used.
sudo docker run --runtime nvidia -it --rm --network host --volume ~/org/nvdli-data:/nvdli-nano/data --device /dev/video0 nvcr.io/nvidia/dli/dli-nano-ai:v2.0.1-r32.5.0

of course its the dli-nano-ai:v2.0.1-r32.5.0 docker image which works on my nano2g just fine. since im using the nano 2g jetpack 4.5.1 os image
but then when i use the xavier nx os image with the same docker image for the DLI the camera has problems.

my xavier with dli-nano-ai:v2.0.1-r32.5.0 container image works now. not sure what i did. i chronicled it at the link mentioned above

OK great. Yes, there is not anything actually specific to Nano about these containers, so it should work on any Jetson.

I have seen that people are talking about the Xavier NX.
My Xavier is the AGX Xavier.
What version of the container should I download?

Hi @aliaga, you can use same container. However the MIPI CSI cameras (like the Raspberry Pi Camera Module v2) that are referred to in the DLI course aren’t able to plug into AGX Xavier’s camera header, because AGX Xavier uses a different camera header than the Nano/NX devkits do. So you can either use a different MIPI CSI camera that you already have working with AGX Xavier, or use a USB webcam (like Logitech C920/C270)

Xavier Nx “usb cam” here and have attempted this a few times over the last month with no success.
Sorry don’t recall what docker did not like at this point in time but i never got past downloading the docker as the image would not start. In fact I have not had any luck with docker on either the Xavier NX or the Nano.

If the docker image fails to start, double-check that you downloaded a container that matches your L4T version. You can check your L4T version with cat /etc/nv_tegra_release

For example, if your L4T version is r32.5.0, then you should download r32.5.0 tag of the container.
The exception is L4T r32.5.1 - that can run the r32.5.0 tag of container.