sisleyli commented on issue #11867: URL: https://github.com/apache/tvm/issues/11867#issuecomment-1165293216
> Actually, the order in the C++ implementation is done intentionally, see #9558 > > cc @AndrewZhaoLuo > This fixes this by moving the clipping operation to operate on the floating point values before casting to integer. My focus is that `round` should be executed before `scale_data` add `expanded_output_zero_point`, right? It seems that no conflict with #9558 , which move clip before cast to integer. -- 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]
