vfdff opened a new issue, #17692:
URL: https://github.com/apache/tvm/issues/17692

   ### Expected behavior
   
   When I repeatedly try to run **matmul_ansor.py** , we 'll get different 
schedule results every time showed in MeasureResult **No: 1 ~ No: 10**,  so I 
expected the following **TIR dump** will matched the above best schedule 
result, .i.e I expect they are different.
   
[matmul_ansor.py.txt](https://github.com/user-attachments/files/19020167/matmul_ansor.py.txt)
   
   ### Actual behavior
   
   The **Lowered TIR** always get same result on my Arm64 target (), ie: dump 
with **print(tvm.lower(sch, args, simple_mode=True))**
   
   ### Environment (172)
   
   Any environment details, such as: Operating System, TVM version
   
   > (/py311-tts) root@998ee80b761b:/home/zhongyunde/source/test/ansor# python 
-c "import tvm; print(tvm.__version__)"
   0.19.dev0
   > (/py311-tts) root@998ee80b761b:/home/zhongyunde/source/test/ansor# cat 
/etc/os-release
   PRETTY_NAME="Ubuntu 22.04.1 LTS"
   
   ### Steps to reproduce
   
   The script is get from https://tvm.hyper.ai/docs/tutorial/ops_AutoScheduling
   then repeatedly run: **python matmul_ansor.py**
   
   For example, I get the 2 logs and compare them, we'll see the part of 
**Lowered TIR** is same, but the prior schedule results are different.
   
![Image](https://github.com/user-attachments/assets/f6c1a268-579e-4f17-90d9-80004a5d4ad6)
   
   ### Triage
   
   Please refer to the list of label tags 
[here](https://github.com/apache/tvm/wiki/Issue-Triage-Labels) to find the 
relevant tags and add them below in a bullet format (example below).
   
   * needs-triage
   


-- 
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]

Reply via email to