javanna merged PR #12270:
URL: https://github.com/apache/lucene/pull/12270
--
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.apa
javanna commented on PR #12270:
URL: https://github.com/apache/lucene/pull/12270#issuecomment-1539665901
Thanks @original-brownbear !
--
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
javanna closed pull request #769: LUCENE-10486: Avoid unnecessary overhead in
TopScoreDoc and TopField collector manager
URL: https://github.com/apache/lucene/pull/769
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
U
javanna commented on PR #12274:
URL: https://github.com/apache/lucene/pull/12274#issuecomment-1539773484
Thanks @iverase !
--
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.
T
javanna merged PR #12274:
URL: https://github.com/apache/lucene/pull/12274
--
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.apa
javanna merged PR #12272:
URL: https://github.com/apache/lucene/pull/12272
--
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.apa
javanna commented on PR #12272:
URL: https://github.com/apache/lucene/pull/12272#issuecomment-1539774583
Thanks @mkhludnev & @iverase !
--
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 specifi
javanna merged PR #12271:
URL: https://github.com/apache/lucene/pull/12271
--
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.apa
javanna commented on PR #12271:
URL: https://github.com/apache/lucene/pull/12271#issuecomment-1539775229
Thanks @iverase !
--
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.
T
rmuir commented on PR #12280:
URL: https://github.com/apache/lucene/pull/12280#issuecomment-1539972391
Please lets not go this path. It is a mutlitermquery, if you want to change
how it works behind the scenes, you "plugin" with RewriteMethod
--
This is an automated message from the Apach
mikemccand commented on issue #12276:
URL: https://github.com/apache/lucene/issues/12276#issuecomment-1540061997
> StringSetAutomatonBuilder?
Hmm can we remove `Set`? (It could be a list or array of String too).
Maybe `StringsToAutomaton`?
--
This is an automated message from the
dweiss commented on issue #12276:
URL: https://github.com/apache/lucene/issues/12276#issuecomment-1540137383
A list or an array - doesn't matter, conceptually it's a set in the end (in
the automaton). But I don't mind any version.
--
This is an automated message from the Apache Git Servic
mikemccand commented on PR #12255:
URL: https://github.com/apache/lucene/pull/12255#issuecomment-1540238456
Hello, is it expected that this change alters the KNN hits returned? That's
fine (if it was expected) ... Lucene's nightly benchmarks are angry about it
though, so I'll just regold i
benwtrent commented on PR #12197:
URL: https://github.com/apache/lucene/pull/12197#issuecomment-1540274950
Thanks for the review @uschindler! I added a test that verifies override
defaults for both methods.
--
This is an automated message from the Apache Git Service.
To respond to the mes
jbellis commented on PR #12255:
URL: https://github.com/apache/lucene/pull/12255#issuecomment-1540274801
Not expected. How can I run the test locally?
--
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 g
msokolov commented on PR #12255:
URL: https://github.com/apache/lucene/pull/12255#issuecomment-1540435331
Those tests are run using a separate package called `luceneutil` -- see
https://github.com/mikemccand/luceneutil
--
This is an automated message from the Apache Git Service.
To respon
msokolov commented on PR #12255:
URL: https://github.com/apache/lucene/pull/12255#issuecomment-1540438392
I wonder if it could have been https://github.com/apache/lucene/pull/12248
that caused the difference?
--
This is an automated message from the Apache Git Service.
To respond to the m
gsmiller commented on PR #12280:
URL: https://github.com/apache/lucene/pull/12280#issuecomment-1540466104
Thanks @rmuir. It would be ideal if we could do this through RewriteMethod,
but I'm not sure how we can actually accomplish that. The problem is in the
implementation of `getTerms`, as
alessandrobenedetti commented on issue #11507:
URL: https://github.com/apache/lucene/issues/11507#issuecomment-1540470731
Hi @dsmiley I updated the dev discussion on the mailing list:
[Proposal] Remove max number of dimensions for KNN vectors
And proceeded with a pragmatic new mail
jbellis commented on PR #12255:
URL: https://github.com/apache/lucene/pull/12255#issuecomment-1540480071
Downloading now. In the meantime, can you give more details on the failure?
Some variance is expected just by the nature of hnsw randomness, but it
shouldn't go from e.g. 90% recall to
zhaih commented on PR #12257:
URL: https://github.com/apache/lucene/pull/12257#issuecomment-1540499897
Thanks @alessandrobenedetti , I'll wait a day to give @msokolov and @jimczi
a chance to review before merging it!
--
This is an automated message from the Apache Git Service.
To respond
gsmiller commented on code in PR #12280:
URL: https://github.com/apache/lucene/pull/12280#discussion_r1188857287
##
lucene/sandbox/src/java/org/apache/lucene/sandbox/search/PKTermInSetQuery.java:
##
@@ -0,0 +1,117 @@
+package org.apache.lucene.sandbox.search;
+
+import java.io.I
rmuir commented on PR #12280:
URL: https://github.com/apache/lucene/pull/12280#issuecomment-1540521962
I still think it doesn't make sense to me to expose this. As i said on the
dev list, your problem is that you use a custom postings format and you want it
to accelerate the intersection.
jimczi commented on PR #12257:
URL: https://github.com/apache/lucene/pull/12257#issuecomment-1540524526
Sorry but that doesn't make sense to me to work/review this PR while
https://github.com/apache/lucene/pull/12254 is under review. I also disagree
with the goal here, the mutable hnsw grap
stefanvodita commented on issue #11547:
URL: https://github.com/apache/lucene/issues/11547#issuecomment-1540554539
I’m curious about this suggestion and I’d like to understand it better. How
could `ConjunctionUtils` tell if an iterator was going to be an “all” iterator?
It couldn’t use cost
zhaih commented on PR #12257:
URL: https://github.com/apache/lucene/pull/12257#issuecomment-1540613109
Sorry @jimczi I am not sure I get your idea.
Did you mean:
1. We shouldn't make improvement to this OnHeapHnswGraph right now because
the other PR is ongoing and it will likely replac
jimczi commented on PR #12257:
URL: https://github.com/apache/lucene/pull/12257#issuecomment-1540666019
I think both 1 and 2 yes. If you think this PR will benefit the other PR
then it should be discussed in https://github.com/apache/lucene/pull/12254. We
already have issues with indexing s
mikemccand commented on PR #12255:
URL: https://github.com/apache/lucene/pull/12255#issuecomment-1540672865
Thanks @jbellis -- I'm not sure this change caused any difference.
Something in the past couple days tweaked the KNN results and this one jumped
out at me as a possibility.
Th
gsmiller closed pull request #12280: Expose iterator over query terms in
TermInSetQuery
URL: https://github.com/apache/lucene/pull/12280
--
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
gsmiller commented on PR #12280:
URL: https://github.com/apache/lucene/pull/12280#issuecomment-1540768011
Got it, thanks @rmuir. I hadn't seen your dev list reply yet. This all makes
sense. I'll close this out and have a look at leveraging intersect. Seems like
a better path forward. Thanks
msokolov commented on PR #12255:
URL: https://github.com/apache/lucene/pull/12255#issuecomment-1540770665
The idea is that the hnsw randomness should be predictable based on its
fixed random seed (42 IIRC). It isn't really a problem if we changed that as
long as we have some idea how / why,
zhaih commented on PR #12257:
URL: https://github.com/apache/lucene/pull/12257#issuecomment-1540937393
I'm ok to just pause it here or eventually drop this PR as I'm definitely
not the one who need this feature (at least for now). But I guess if @jimczi
you want to stop people from using th
jbellis commented on PR #12255:
URL: https://github.com/apache/lucene/pull/12255#issuecomment-1541171234
If the algorithm is implemented correctly, and I think that it is, then in
theory the order of neighbor traversal should not matter.
But we are seeing a difference here, so I *thin
ryantbrown commented on issue #11507:
URL: https://github.com/apache/lucene/issues/11507#issuecomment-1541351180
The rabbit hole that is trying to store Open AI embeddings in Elasticsearch
eventually leads here. I read the entire thread and unless I am missing
something, the obvious move t
34 matches
Mail list logo
