cccclai
cccclai
Can I get a review on this PR?
As a note, this pr just test fp, and quantized version need to be added later
> I think it's more intuitive to make it part of the existing `executorch-ubuntu-22.04-clang12-android` docker image, otherwise have to switch to load different dockers to build QNN runtime and android...
Closing stale pr
@pytorchbot cherry-pick --onto release/0.4 -c docs
Did you run these two commands? ``` cp -f backends/qualcomm/PyQnnManagerAdaptor.cpython-310-x86_64-linux-gnu.so $EXECUTORCH_ROOT/backends/qualcomm/python cp -f backends/qualcomm/PyQnnWrapperAdaptor.cpython-310-x86_64-linux-gnu.so $EXECUTORCH_ROOT/backends/qualcomm/python ``` Full instruction here https://pytorch.org/executorch/main/build-run-qualcomm-ai-engine-direct-backend.html
cc: @digantdesai, @mcr229
The error shows up when running this script `https://github.com/pytorch/executorch/blob/main/backends/qualcomm/scripts/build.sh` based on the log. If you have a linux machine, can you follow https://pytorch.org/executorch/stable/build-run-qualcomm-ai-engine-direct-backend.html and see if the script fails?
> @cccclai I finally managed to reproduce the issue by running script `backends/qualcomm/scripts/build.sh` with the parameters from the CI script [here](https://github.com/pytorch/executorch/blob/main/.ci/scripts/build-qnn-sdk.sh#L17). Interestingly, the issue seems to be caused by `--job_number...
I think the buck version should be set correctly if you install executorch. MTK backend didn't use speicalized buck, but the same as the executorch buck