[GitHub] [tvm] MrJungle1 commented on issue #15342: [Bug] The optimal implementation of reduce_sum searched by Ansor is more than 30x slower than torch.sum

[via GitHub]

MrJungle1 commented on issue #15342:
URL: https://github.com/apache/tvm/issues/15342#issuecomment-1641951831

   In addition, in another case, shape = (120, 128, 128, 128) axis = (1,2) 
generated very common code
   
![image](https://github.com/apache/tvm/assets/43968255/7fab2aad-0d2a-4b20-8909-788bb604ca24)
   In addition, in another case, shape = (128, 8, 128, 128) axis = (1,)
   
![image](https://github.com/apache/tvm/assets/43968255/6ccbb509-425e-47d7-a7b8-97163ad34094)
   
   
   


-- 
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: commits-unsubscr...@tvm.apache.org

For queries about this service, please contact Infrastructure at:
us...@infra.apache.org