slyubomirsky commented on PR #14324: URL: https://github.com/apache/tvm/pull/14324#issuecomment-1474094980
It would be nice to have a single source of truth for shape inference, if that's what you're referring to in question 2. What exactly is question 1 referring to? You can use `MatchCast` to bind the inferred dimensions to new variables to avoid having to copy over cumbersome expressions, but we should be able to nest PrimExprs in symbolic shapes and do further algebraic reasoning on them -- 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]
