mikemccand commented on PR #12526:
URL: https://github.com/apache/lucene/pull/12526#issuecomment-1700726359
Wow, impressive! Maybe we should add `OrHighVeryLow` to nightly benchy too?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to
mikemccand commented on issue #7350:
URL: https://github.com/apache/lucene/issues/7350#issuecomment-1700812722
`FieldCache` is indeed long gone from Lucene -- I'm not sure how we missed
adding this to `MIGRATE.txt`.
For efficient range filtering, it is usually best now to index your n
mikemccand commented on issue #7047:
URL: https://github.com/apache/lucene/issues/7047#issuecomment-1700832220
@PenghaiZhang sorry about all this.
Given how ancient these Lucene versions are, few people even remember the
specifics of how to do this migration. Maybe you could jot some
mikemccand commented on issue #12527:
URL: https://github.com/apache/lucene/issues/12527#issuecomment-1700871714
I like this idea, reducing possible IO overhead. But I tested it with
`luceneutil` on `wikimediumall`:
```
TaskQPS base StdDev QP
benwtrent commented on PR #12421:
URL: https://github.com/apache/lucene/pull/12421#issuecomment-1700939168
Haven't forgotten about this. Just been bogged down with other things. Hope
to revisit again soon!
--
This is an automated message from the Apache Git Service.
To respond to the mess
mikemccand opened a new pull request, #12530:
URL: https://github.com/apache/lucene/pull/12530
### Description
Relates #7820.
`CheckIndex` today only detects and exorcises corruption with the latest
commit point, yet `IndexWriter` will be angry on init if there are older commit
mikemccand commented on issue #7820:
URL: https://github.com/apache/lucene/issues/7820#issuecomment-1700986388
> Third off, there is possibly a separate improvement we could make to
IndexWriter, to remove segments_N files before removing all other files when a
commit point is deleted, to tr
mikemccand commented on PR #12530:
URL: https://github.com/apache/lucene/pull/12530#issuecomment-1701119507
Thanks @rmuir
> as far as the exorcise stuff, I think a good next step would be to start
writing some unit tests that invoke exorcise? we have a grand total of zero
tests exer
benwtrent commented on code in PR #12529:
URL: https://github.com/apache/lucene/pull/12529#discussion_r1311719842
##
lucene/backward-codecs/src/test/org/apache/lucene/backward_codecs/lucene94/Lucene94HnswVectorsWriter.java:
##
@@ -630,7 +621,8 @@ private abstract static class Fi
jimczi commented on code in PR #12529:
URL: https://github.com/apache/lucene/pull/12529#discussion_r1311935419
##
lucene/core/src/java/org/apache/lucene/util/hnsw/HnswGraphSearcher.java:
##
@@ -172,106 +83,36 @@ public static KnnCollector search(
* @return a set of collected
Jeevananthan-23 opened a new issue, #12531:
URL: https://github.com/apache/lucene/issues/12531
### Description
In **LuceneNet**(`C#`) we had a long conversation about adding support for
[Async API issue](https://github.com/apache/lucenenet/issues/763) but the scope
of the project was
benwtrent commented on PR #12529:
URL: https://github.com/apache/lucene/pull/12529#issuecomment-1701731417
Ran a benchmark on Lucene util and here are the results, candidate (this PR)
is consistently slightly slower.
```
recall latency nDocfanout maxConn beamWidth visit
zhaih commented on code in PR #12480:
URL: https://github.com/apache/lucene/pull/12480#discussion_r1312372706
##
lucene/CHANGES.txt:
##
@@ -90,6 +90,8 @@ Optimizations
* GITHUB#12408: Lazy initialization improvements for Facets implementations
when there are segments with no
Tony-X commented on issue #12513:
URL: https://github.com/apache/lucene/issues/12513#issuecomment-1701907189
I'd like to seek for some advices regarding the situation I am in --
I want to preserve the nice properties of the tantivy's termdict as I port
it over for Lucene
1. defini
easyice opened a new pull request, #12532:
URL: https://github.com/apache/lucene/pull/12532
### Description
Sice https://github.com/apache/lucene-solr/pull/1464 , the default for
maxPointsPerLeafNode is changed from 1024 to 512, some comments is outdated.
--
This is an
zhaih commented on PR #12480:
URL: https://github.com/apache/lucene/pull/12480#issuecomment-1702214309
@Jackyrie2 Seems the precommit fails?
--
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 sp
zhaih opened a new issue, #12533:
URL: https://github.com/apache/lucene/issues/12533
### Description
Currently when we're merging HNSW graphs we're able to start with an exiting
graph but not inserting nodes from scratch thanks to #12050. But we have set a
constraint that the init gr
zhaih commented on issue #12440:
URL: https://github.com/apache/lucene/issues/12440#issuecomment-1702229556
One problem with optimization #12050 is that with a hybrid text-embedding
index you can hardly have even one big segment that do not have the deletion,
which makes real world index me
zhaih commented on issue #12440:
URL: https://github.com/apache/lucene/issues/12440#issuecomment-1702237913
Another idea I have is a bit wild: what if we do less merge?
For example, if we have segment 1,2,3,4 wants to merge and form a new
segment, can we just leave the HNSW graphs as-
19 matches
Mail list logo
