we have converted a inception model to .dlc to run on a QMC6490. This model is part of examples provided in the snpe sdk.
we are using snpe 2.21.0.240401 to quantize the model.
We are using snpe aar in an android demo app.
we observe memory leak in the native code when we run the app for more than few minutes. see the screenshot below.
We also observe the same behavior when using a custom yolo model.
What could we do to get this addressed?!
Opinions expressed in the content posted here are the personal opinions of the original authors, and do not necessarily reflect those of Qualcomm Incorporated or its subsidiaries (“Qualcomm”). The content is provided for informational purposes only and is not meant to be an endorsement or representation by Qualcomm or any other party. This site may also provide links or references to non-Qualcomm sites and resources. Qualcomm makes no representations, warranties, or other commitments whatsoever about any non-Qualcomm sites or third-party resources that may be referenced, accessible from, or linked to this site.
We now have run the snpe provided app that runs in a loop to observe memory growth. we observe exponential memory growth after few hours. any ideas to overcome this?