-
Notifications
You must be signed in to change notification settings - Fork 31
Speed-up time-based samplers by 20X and index-based by 1.5X #284
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
Conversation
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Changes made to this file are only to:
- assert the new output type
- slightly udpate some parts to account for the new return type
I didn't add or remove any test
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Tests are passing so this is great.
Non-blocking commentit would be nice to compare frame contents against reference clip contents gathered from the test data files (not comparing clip vs. another) for more confidence.
Thanks for the review! I'll follow-up with the tests you're mentioning soon |
Fixes #256
We now let the samplers rely on our C++ "sort and dedup" logic, instead of the less efficient Python ones. This has a few benefits:
(num_clips, num_frames_per_clips, C, H, W)
(or HWC).Running our samplers benchmark:
🚀