masahi opened a new pull request #5325: [Refactor] Add memoized expr translator for use by backend codegen URL: https://github.com/apache/incubator-tvm/pull/5325 Related discussion: https://discuss.tvm.ai/t/missing-memoization-in-exprfunctor/6334 The rationale is two fold: * Clean up existing code that has duplicated memo logic and unnecessary `VisitExpr(const Expr& n)` override * More importantly, to remove the cognitive load of having to do memoization by external codegen implementers. Most likely any external codegen needs some memoization. It would be better to take care of this in the Relay side, so that there would be no chance of forgetting to do memoization and having a bad time (which happened to me). please review @tqchen @zhiics
---------------------------------------------------------------- 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] With regards, Apache Git Services
