electriclilies commented on pull request #9334: URL: https://github.com/apache/tvm/pull/9334#issuecomment-949003599
@masahi Yeah I think I actually added the TNonComputational option to the QNN ops-- conceptually I guess they are more "symbolic" than non-computational, since they do conceptually compute something, but that compute is injected during legalization, which is wayyy before lowering. I added it because when I was working on quantization, one time I accidentally skipped the legalize pass and then lowering broke in a weird way. But with the TNoncomputational attr set it would just throw an error. -- 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]
