JackWeiw commented on PR #16679: URL: https://github.com/apache/tvm/pull/16679#issuecomment-1985883927
> Hmm, that's the tricky part. This assertion is specific to your use case, but isn't true in general. For symbolic shapes in both TIR and Relax, it is legal for the axis to have an extent of zero. How are you currently generating the PrimFunc from your script? In my use case i would get IRModule generate from mlc-llm, and as i am concerned there are most two simbolic shape m and n, and i can extract symbolic shape infomation from params of func. Maybe i can simplify add a rewrite pass to add T.Assert(m>0) in every primfunc if the are dynamic args, i will try to do it. Please let me know if you have a better insight! -- 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]
