microsoft / microsoft/WindowsAppSDK-Samples

[WinML] Debug break observed on QNN device (ARM64) w/ CppResnetDemo app

Open
#486 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
PowerShell
Stars
1k
Forks
335
Avg merge
5d 1h
Merged PRs (30d)
8

Description

Steps

  1. Clone the repository onto a QNN device
  2. Use the AI Toolkit conversion flow to create a Qualcomm-compatible version of the SqueezeNet model
  3. Update the placeholder paths in the CppResnetDemo application to point to this model, and set a path for the compiled model
  4. F5 the application

Expected
The application runs without issue and the user gets back a set of ranked predictions for the input image

Actual
The application hits a debug break at the CreateTensorWithDataAsOrtValue API call.

(This issue also repro'd for me on the Experimental 1 WinML release, so it is not specific to the WinAppSDK implementation.)

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start with the CppResnetDemo application and reproduce the QNN ARM64 flow using the AI Toolkit-converted SqueezeNet model and the configured model paths. Investigate the debug break at CreateTensorWithDataAsOrtValue; done means the application runs without breaking and returns ranked predictions for the input image.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
desktop, machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.