Open
Description
Prerequisites
- I am running the latest code. Mention the version if possible as well.
- I carefully followed the README.md.
- I searched using keywords relevant to my issue to make sure that I am creating a new issue that is not already open (or closed).
- I reviewed the Discussions, and have a new and useful enhancement to share.
Feature Description
Add op verbose or event to print op, kernel_name，exec_time, shape, algo ...
Motivation
When users or developers use different hardware to test different models, may be performance data that does not match theoretical expectations.
We expect to break down the model, find the bottleneck op, and then optimize the operator.
Kernel level verbose/event is the basis of performance optimization.
Possible Implementation
No response
Activity