jwfromm opened a new pull request, #16427: URL: https://github.com/apache/tvm/pull/16427
I've recently been working with dynamic shaped convolutions and found a few blockers that this PR resolves. I enable the NN frontend to work with dynamic axes and also introduce a small check to cutlass pattern matching to skip convolutions with dynamism rather than error out. I add appropriate tests to make sure the behaviors work as expected. -- 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]
