Is there a version of jetpack preinstalled with jupyterlab with TF2.X?

Is there a version of jetpack preinstalled with jupyterlab with TF2.X?

Hi,

Unfortunately, we don’t have a container with both installed.

You can find a JupyterLab+TensorFlow 1.15.x container in l4t-ml.
Or a TensorFlow 2.x package in l4t-tensorflow.

An simple alternative is to upgrade the TensorFlow version in l4t-ml.
Below is a command for JetPack4.6:

$ apt-get update
$ pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v46 tensorflow==2.6.0

Thanks.

I tried the docker container. Works fine with TF1.15.

But as suggested when I try upgrading it to TF2 I am having these errors. Attached doc. TF2 issue with Jetson nano 4GB.docx (893.6 KB)

How can I solve this?

Thank you

Hi,

Thanks for reporting this.

We are checking this issue internally.
Will share more information with you later.

Hi,

Since our v2.6.0+nv21.9 TensorFlow release has some issues, please use v2.5.0+nv21.8 instead.

For example, we can setup a container with both jupyterlab and TF v2.5 via the following commands:

$ sudo docker run -it --rm --runtime nvidia --network host nvcr.io/nvidia/l4t-ml:r32.6.1-py3
> apt update
> pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v46 tensorflow==2.5.0+nv21.8
$ pip3 install --pre --extra-index-url https://developer.download.nvidia.com/compute/redist/jp/v46 tensorflow==2.5.0+nv21.8
Collecting tensorflow==2.5.0+nv21.8
  Downloading https://developer.download.nvidia.com/compute/redist/jp/v46/tensorflow/tensorflow-2.5.0+nv21.8-cp36-cp36m-linux_aarch64.whl (313.8MB)
    100% |################################| 313.8MB 3.7kB/s
...
Successfully installed absl-py-0.12.0 cachetools-4.2.4 flatbuffers-1.12 gast-0.4.0 google-auth-2.3.0 google-auth-oauthlib-0.4.6 grpcio-1.34.1 keras-nightly-2.5.0.dev2021032900 keras-preprocessing-1.1.2 numpy-1.19.5 oauthlib-3.1.1 pyasn1-0.4.8 pyasn1-modules-0.2.8 requests-oauthlib-1.3.0 rsa-4.7.2 six-1.15.0 tensorboard-2.7.0 tensorboard-data-server-0.6.1 tensorboard-plugin-wit-1.8.0 tensorflow-2.5.0+nv21.8 tensorflow-estimator-2.5.0 typing-extensions-3.7.4.3

Thanks.

Hello AastaLLL,

I was able to install TF2.5 with your instruction. When I run it first time from the browser (IP:8888) I am able to log in to jupyter notebook. But once I shut down the system and reboot it I am not able to access jupyter notebook any more. I have tried this process with several wifi connections and I tried with uninstalling docker few times. Every time I am able to access jupyter notebeook but once I restart no more access. I can SSH to the jetson though so I know my wifi is working fine. I did not have this with the normal docker installation. Once I update to your instructed version I am starting to have this access issues. What do you think the issue is?

Thank you

Hello AastaLLL,
I was wondering if you are able to suggest me a solution for my earlier post ?

Thank you

Hi,

We test this in our environment but the access still works after rebooting.
Would you mind manually restarting the service to see if it helps?

$ jupyter lab --ip 0.0.0.0 --port 8888 --allow-root &> /var/log/jupyter.log

Thanks.

This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.