I tried running the pytorch mnist example on a train-agent by running it locally and then resetting the experiment and then enqueue-ing it to the default queue.
All works well but it seems the environment building process gets stuck on a manual flag triggered by an install (tzdata). Attached is a screen shot of the log.
Looking online brought up the following link ( https://rtfm.co.ua/en/docker-configure-tzdata-and-timezone-during-build/ ). It might be a problem with the docker that the agent builds for the experiment (if i understood how the agent works correctly). How do i go about resolving this?
Thanks again for all you help.