dungba88 commented on issue #13564: URL: https://github.com/apache/lucene/issues/13564#issuecomment-2481270142
> Hey @dungba88 - not intrusive at all! I ended up not having more time to work more on this. I had a [simple PoC](https://github.com/dungba88/lucene/pull/29/) to demonstrate this idea (test has passed as well). I created a separate Query for sake of simplicity, but also proves that Lucene already enable this idea with (almost) no change needed in core. The idea is to add a `float oversample` parameter in constructor, which will sample more matches and run the second pass for FP reranking. Although the second pass currently has to be done at each segment level, as I still don't know how to get the raw FP vectors without the LeafReaderContext, so I can't compute vector similarity with just the TopDocs alone. -- 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: issues-unsubscr...@lucene.apache.org For queries about this service, please contact Infrastructure at: us...@infra.apache.org --------------------------------------------------------------------- To unsubscribe, e-mail: issues-unsubscr...@lucene.apache.org For additional commands, e-mail: issues-h...@lucene.apache.org
