kaivalnp commented on issue #14758:
URL: https://github.com/apache/lucene/issues/14758#issuecomment-3294554802
> What if it only did so within one document, which would enable this
"compile KNN prefilter to separate field's HNSW graph during indexing"
efficiently? But not across documents
rmuir closed issue #14556: Smoke tester requiring Python 3.12+
URL: https://github.com/apache/lucene/issues/14556
--
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
github-actions[bot] commented on PR #15134:
URL: https://github.com/apache/lucene/pull/15134#issuecomment-3294419368
This PR has not had activity in the past 2 weeks, labeling it as stale. If
the PR is waiting for review, notify the d...@lucene.apache.org list. Thank you
for your contributi
mccullocht opened a new pull request, #15183:
URL: https://github.com/apache/lucene/pull/15183
This was introduced in #15021 where we don't correctly account the max score
when there are 3 remaining
vectors in a bulk scoring call.
Fixed #15180 15180
--
This is an automated messa
mccullocht commented on PR #15169:
URL: https://github.com/apache/lucene/pull/15169#issuecomment-3294250569
Average visited count in the query path actually is exposed in luceneutil
today, it just appears in the iteration summary and not the overall summary.
TIL. I've extracted it for my mo
rmuir merged PR #15162:
URL: https://github.com/apache/lucene/pull/15162
--
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.apach
benwtrent commented on PR #15169:
URL: https://github.com/apache/lucene/pull/15169#issuecomment-3294068941
> Marked the old codecs as deprecated. I'd prefer not to do backwards codecs
here, this change is already much larger than I'd like but I couldn't figure
out how to factor it into smal
mccullocht commented on code in PR #15169:
URL: https://github.com/apache/lucene/pull/15169#discussion_r2350130404
##
lucene/core/src/java/org/apache/lucene/codecs/lucene104/Lucene104ScalarQuantizedVectorsFormat.java:
##
@@ -0,0 +1,206 @@
+/*
+ * Licensed to the Apache Software
msokolov commented on PR #15090:
URL: https://github.com/apache/lucene/pull/15090#issuecomment-3292182686
It reminds me of something I learned about at ApacheCoC where @kaivalnp
presented the Faiss vectors format -- FAISS expects its users to provide a
complex string describing the format i
rmuir commented on issue #15185:
URL: https://github.com/apache/lucene/issues/15185#issuecomment-3293617192
I agree, I tried to make some progress the other day and seemed to only
uncover more breaks and problems.
--
This is an automated message from the Apache Git Service.
To respon
dweiss commented on issue #15185:
URL: https://github.com/apache/lucene/issues/15185#issuecomment-3293650750
I've removed it completely on my dev branch. Will reimplement basic license
checks tomorrow. I think this should be simpler to manage and comprehend.
--
This is an automated messag
dweiss commented on issue #15185:
URL: https://github.com/apache/lucene/issues/15185#issuecomment-3293517589
Related - https://github.com/apache/lucene/pull/14582
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL
dweiss opened a new issue, #15185:
URL: https://github.com/apache/lucene/issues/15185
This is a quick check if anybody minds...
I've had a few attempts to refactor apache rat code (that checks file
licenses) and I have to say my conclusion is that it'd be much simpler to just
write a
dweiss commented on PR #15162:
URL: https://github.com/apache/lucene/pull/15162#issuecomment-3293482243
Yes, it's fine with me - feel free to merge, I'll follow-up with workflows
changes.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on
dweiss commented on PR #15162:
URL: https://github.com/apache/lucene/pull/15162#issuecomment-3293261437
Ok, nevermind - I think it's fine even if we don't use venv there in that
workflow, the container should have an up to date version of all dependencies.
I'll follow up if it breaks.
--
zhaih commented on code in PR #15184:
URL: https://github.com/apache/lucene/pull/15184#discussion_r2349546069
##
lucene/core/src/java/org/apache/lucene/util/hnsw/HnswGraphBuilder.java:
##
@@ -144,8 +148,8 @@ protected HnswGraphBuilder(
throw new IllegalArgumentException("
javanna commented on issue #13745:
URL: https://github.com/apache/lucene/issues/13745#issuecomment-3291991178
@smuching202 the two I remember clearly about are `PointRangeQuery` and
`PointInSetQuery`. These build a bitset ahead of time of the size of the entire
segments. There may be others
msokolov commented on PR #15090:
URL: https://github.com/apache/lucene/pull/15090#issuecomment-3292278407
> Are we cool with writing that stuff to metadata and loading? If so, I can
make a chnage this week so we can see it in action and then go through the
process of adding a new HNSW forma
benwtrent commented on PR #15090:
URL: https://github.com/apache/lucene/pull/15090#issuecomment-3292210203
> I guess people working in Java are here at least in part for the
type-safety and compile-time checking, but perhpas we could create a format -
builder? Java developers love builders
msokolov commented on PR #15090:
URL: https://github.com/apache/lucene/pull/15090#issuecomment-3292199372
I mean at the end of the day we need to have a format name that corresponds
to a class name so that SPI can load the format, but the format itself can have
some parameterization that is
msokolov commented on code in PR #15184:
URL: https://github.com/apache/lucene/pull/15184#discussion_r2348953441
##
lucene/core/src/java/org/apache/lucene/util/hnsw/HnswGraphBuilder.java:
##
@@ -144,8 +148,8 @@ protected HnswGraphBuilder(
throw new IllegalArgumentExceptio
21 matches
Mail list logo
