Hi, Is it possible to submit a training model to use a portion of preprocessed data when the preprocessing is still underway? I have submitted a preprocessing only docker but now I want to submit a training docker with the same preprocessing docker in the docker_id.txt hoping to run the training on whatever is available. Would it work? I don't want to restart the preprocessing and erase what I have gathered so far. Thanks, Michael

Created by Michael Hsieh michaelhsieh
> Is it possible to submit a training model to use a portion of preprocessed data when the preprocessing is still underway? Yes, you can *submit* a training submission that refers to a pre-processing only image that has been previously submitted. Your training submission will only start running once your pre-processing only submission has been completed. If the repository name + digest of the pre-processing Docker image is the same in both your pre-processing only and training submissions, then the system will not run the pre-processing container a second time (assuming that there are no other reasons for running again your pre-processing step). Please let me know if this doesn't answer your question. Thanks!

Start the training before the preprocessing completes page is loading…