How to rebuild the docker image based on ros deep learning?

Hi everyone! I’m currently working on jetbot_ros image which has built in ros_deep_learning node I have to download the pretrained model whenever i rerun that docker container. Is there any for me to save that progress or is there any docker image that has already downloaded pretrianed model. Thanks in advance

Hi,

The Dockerfile for jetbot_ros can be found below:
You can modify the file and build a custom image with model pre-download.

Thanks.

You should also be able to mount the /jetson-inference/data directory to your device so that it saves the model(s) outside container. To do this, first clone the jetson-inference repo to your device (outside container)

Then add this line to your jetbot_ros/docker/run.sh script around line 195:
https://github.com/dusty-nv/jetbot_ros/blob/e2bd28b3e182d772ef051f1cc0e5c42c9fb4f1ec/docker/run.sh#L195

--volume /path/to/your/jetson-inference/data:/jetson-inference/data

Substitute the path to the jetson-inference repo that you cloned on your device in the line above.

Thanks for all your reply!

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