mboehm7 commented on PR #2430: URL: https://github.com/apache/systemds/pull/2430#issuecomment-3919417885
@kubraaksux Thanks for the awesome contribution. I think the current approach of calling the dedicated python server is a great fallback and baseline. Ultimately, we want to convert such pretrained models to SystemDS matrices and run the inference natively with SystemDS operations. However, lots of primitives and performance optimizations are still missing, so this baseline is a great start. Before merging this PR, I would recommend to also have some negative tests that check for robustness, e.g., what happens if the python server was accidentially killed. -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
