szha commented on pull request #19209: URL: https://github.com/apache/incubator-mxnet/pull/19209#issuecomment-698556256
That's a good point. It's unclear to me what the performance impact would be if we don't allow multiple copies of random states in case one doesn't care about reproducibility. Also, I'm not sure whether waitall resolves the race condition completely in the case where multi-threaded execution happens and we use the thread-safe cached op. ---------------------------------------------------------------- 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. For queries about this service, please contact Infrastructure at: [email protected]
