I'm running the main pipeline code remotely(not using run_locally).
The component in question seems to fail almost immediately after pipeline execution because of value error.
If I put docker argument in the decorator, value error appears and becomes failure..
A follow up, (1) how are you spinning the agent ? (2) could it be the docker image "ultralytics/yolov5" does not have Bash as entry point ?
you can force that with
@PipelineDecorator.component(return_values=['int'], cache=False, task_type='training', docker="ultralytics/yolov5", docker_args="--entrypoint /bin/bash", parents=['preprocessing'])
Just to make sure, the first two steps are working ?
Maybe it has to do with the fact the "training" step specifies a docker image, could you try to remove it and check?
BTW: A few pointers
return_values is used to specify multiple returned objects stored individually, not the type of the object. If there is a single object, no need to specify
parents argument is optional, the pipeline components optimizes execution based on inputs, for example in your code, all pipeline components will be running in parallel, because they do not rely on the return value of one another, unless of course if you specify "
parents " 🙂