-
Notifications
You must be signed in to change notification settings - Fork 2.3k
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 our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Optimize text_classifer model memory usage #21960
Comments
For QA:
|
Verification passed on
Reproduced the issue in 1.38.xVerify that the issuers were downloaded
Verify that unblinded tokens are downloaded
Verified ad related components are downloaded in NTP Sponsored Images (US) - Version: 1.0.803 Viewed an ad and dimissed it The memory used is 37352KBVerified fixed in 1.39.xVerify that the issuers were downloaded
Verify that unblinded tokens are downloaded
Verified ad related components are downloaded in NTP Sponsored Images (US) - Version: 1.0.803 Viewed an ad and dimissed it The memory used is 19252KB, which is less than in 1.38.x |
@LaurenWags @LaurenWags @btlechowski Big thanks for verification! Just want to clarify one thing: |
Makes sense, thanks for the clarification @atuchin-m 👍🏻 |
Verification PASSED on
Reproduced the issue in 1.38.xVerify that the issuers were downloaded
Verify that unblinded tokens are downloaded
Verified ad related components are downloaded in NTP Sponsored Images (CA) - Version: 1.0.149 Viewed an ad and dismissed it The memory used is
|
Verification PASSED on
Reproduced the issue in 1.38.xVerify that the issuers were downloaded
Verify that unblinded tokens are downloaded
Verified ad related components are downloaded in NTP Sponsored Images (CA) - Version: 1.0.149 Viewed an ad and dismissed it The memory used is
|
Related issue: #21602
Parsed text classifier model (component id :
emgmepnebbddgnkhfmhdhmjifkglkamo
) takes ~24mb of memory in Ads Service process.It's looks excessive and we should find a way to make it smaller.
Another problem is that peak usage on start is more than 200mb.
Here is the code that produce 24mb memory allocation: https://github.com/brave/brave-core/blob/master/vendor/bat-native-ads/src/bat/ads/internal/ml/pipeline/text_processing/text_processing.cc#L63
The text was updated successfully, but these errors were encountered: