jiangjiajun commented on issue #11435: URL: https://github.com/apache/tvm/issues/11435#issuecomment-1136679142
I think the difference may comes from the comparison between cpu and gpu. Since all the computation of PaddlePaddle is running on the cpu device, and tvm will run both on cpu and gpu. @heliqi Could you help to send a pull request to raise the tolerance to 1e-04? -- 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]
