Pulkitg64 commented on issue #16906: URL: https://github.com/apache/lucene/issues/16906#issuecomment-6533640865
Thanks @Tony-X. This looks interesting and results looks promising as also confirmed by @xande. While I am still going through the code changes, I had some few basic questions: > A test about the parallel potential Do you know why are we seeing this speed up? Is it because of the bulk scoring we are doing with this new approach or number of similarity computation are also reduced. I mean do you have data around how many score computation in level 0 we are doing with new approach? Also I am assuming you are using fp32 vectors for these tests. Do you think it is worth doing some test with quantized vectors as well? -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected] --------------------------------------------------------------------- To unsubscribe, e-mail: [email protected] For additional commands, e-mail: [email protected]
