Qualcomm AI Engine Direct - UT fix#18787
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/18787
Note: Links to docs will display an error until the docs builds have been completed. ✅ You can merge normally! (2 Unrelated Failures)As of commit f584c35 with merge base 137dedc ( BROKEN TRUNK - The following jobs failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This PR needs a
|
f8f6733 to
f584c35
Compare
Summary
Currently we run UT like the following:
python backends/qualcomm/tests/test_qnn_delegate.py -k TestQNNQuantizedUtils --model SM8750 --device ed8bb642 --build_folder build-androidThis runs all tests under
TestQNNQuantizedUtils. However, if the first test sets attributes likeTestQNN.enable_profile=True, this will get carry over to rest of the tests and all tests will haveenable_profileequals to True, which is not desired. This change takes advantage issetupfunction in UT to reset attributes back toFalsebetween executions.Test plan
Passing all
Utilstests underexecutorch/backends/qualcomm/tests/test_qnn_delegate.py