- Notifications
You must be signed in to change notification settings - Fork3.5k
[QNN EP] Adjust tolerance for Clip and Transpose tests due to FP16 default in QNN HTP#26499
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to ourterms of service andprivacy statement. We’ll occasionally send you account related emails.
Already on GitHub?Sign in to your account
Open
yuhuchua-qti wants to merge1 commit intomicrosoft:mainChoose a base branch
base:main
Could not load branches
Branch not found:{{ refName }}
Loading
Could not load tags
Nothing to show
Loading
Are you sure you want to change the base?
Some commits from the old base branch may be removed from the timeline, and old review comments may become outdated.
Uh oh!
There was an error while loading.Please reload this page.
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.Learn more about bidirectional Unicode characters
* Updated tolerance thresholds to align with FP16 becoming the default precision in QNN HTP.
@microsoft-github-policy-service agree company="Qualcomm" |
Sign up for freeto join this conversation on GitHub. Already have an account?Sign in to comment
Labels
None yet
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Description
This PR updates the tolerance thresholds for the Clip and Transpose tests in QnnHTPBackendTests. The adjustment accounts for minor accuracy differences introduced by the change in default floating-point precision in QNN HTP starting from version 2.35.
Motivation and Context
Since QNN 2.35, the default floating-point precision in QNN HTP has changed from FP32 to FP16. Additionally, the configuration option
QNN_HTP_GRAPH_CONFIG_OPTION_PRECISIONhas been deprecated.This change in precision can lead to expected accuracy loss, especially in scenarios where graph inputs and outputs are defined as float 32, but internal computations are performed in FP16 (e.g., FP32 → FP16 → FP32 conversions). To accommodate this, the tolerance thresholds for the affected tests have been increased to prevent false negatives due to precision differences.
@microsoft-github-policy-service agree company="Qualcomm"