You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Clean up for runner may potentially also include the stopped container (not sure if that already happens), or the image (though would not do that by default).
The text was updated successfully, but these errors were encountered:
Currently, cleanup is not implemented in any derived class, but is always called after a run.
I definitely would not want to remove the docker image after each run.
Perhaps the clean should only be called on --setup modes clean (#491) or force.
Then it could be something like:
defcleanup(self):
image=self._find_image()
log.info(f"Cleaning up docker image {image}.")
run_cmd(f"docker rmi -f {image}")
log.info("Cleaning up Dockerfile")
ifos.path.exists(self._script):
os.remove(self._script)
Moving TODOs from code to issue tracker (#172):
automlbenchmark/amlb/runners/container.py
Line 95 in 98bf554
Clean up for runner may potentially also include the stopped container (not sure if that already happens), or the image (though would not do that by default).
The text was updated successfully, but these errors were encountered: