navneet1v opened a new issue, #13920:
URL: https://github.com/apache/lucene/issues/13920
## Description
For past 1 month we have been testing difference in performance for a files
getting opened with IOContext.RANDOM vs IOContext.READ specially during merges
with Lucene version 9.11.1 an
jpountz commented on PR #13885:
URL: https://github.com/apache/lucene/pull/13885#issuecomment-2416031098
It's almost certainly this change that sped up TermTitleSort on October
12th, since TermTitleSort is the task that creates the most `PostingsEnum`
objects. I pushed an annotation.
--
hanqiushi opened a new issue, #13921:
URL: https://github.com/apache/lucene/issues/13921
### Description
Hi, I have a question when using hnsw int4.
I have built an index using Lucene99HnswScalarQuantizedVectorsFormat bits=4
param. But I don't know how to use it in search.
I hav
stefanvodita commented on PR #13740:
URL: https://github.com/apache/lucene/pull/13740#issuecomment-2416197652
I like progress-not-perfection, but looking at this again I'm not sure it's
progress. To me, it seems like a lot of complexity for a little bit of
efficiency, but I could be wrong.
jpountz commented on code in PR #13872:
URL: https://github.com/apache/lucene/pull/13872#discussion_r1802496621
##
lucene/MIGRATE.md:
##
@@ -892,3 +892,7 @@ segments are rewritten either via `IndexWriter.forceMerge`
or
### Vector values APIs switched to primarily random-access
benwtrent commented on PR #13635:
URL: https://github.com/apache/lucene/pull/13635#issuecomment-2417118146
Hey @seanmacavaney didn't want this to die on the vine. I think with some
refactoring and adding new experimental queries, this could be a nice
experimental feature for vector search.
seanmacavaney commented on PR #13635:
URL: https://github.com/apache/lucene/pull/13635#issuecomment-2417194138
Not at all-- thanks a lot for the help @benwtrent! I totally agree with the
proposed changes and it's clear how to move forward on this. I'm just occupied
with several other priori
bugmakerr commented on PR #13763:
URL: https://github.com/apache/lucene/pull/13763#issuecomment-2416803313
@jpountz I have merged main branch, PTAL:)
--
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
bugmakerr commented on code in PR #13763:
URL: https://github.com/apache/lucene/pull/13763#discussion_r1803076020
##
lucene/core/src/java/org/apache/lucene/codecs/perfield/PerFieldKnnVectorsFormat.java:
##
@@ -239,51 +245,69 @@ public FieldsReader(final SegmentReadState read
jpountz commented on PR #13904:
URL: https://github.com/apache/lucene/pull/13904#issuecomment-2416808261
This caused a big regression in `AndHighOrMedMed`. I'm reverting.
https://benchmarks.mikemccandless.com/AndHighOrMedMed.html
--
This is an automated message from the Apache Git Service
benwtrent commented on code in PR #13919:
URL: https://github.com/apache/lucene/pull/13919#discussion_r1803334241
##
lucene/core/src/java/org/apache/lucene/util/quantization/ScalarQuantizer.java:
##
@@ -568,29 +568,34 @@ private static List
findNearestNeighbors(
* and `95`.
benwtrent opened a new issue, #13922:
URL: https://github.com/apache/lucene/issues/13922
### Description
It would take a bit of refactoring, but:
```
float dx = v - minQuantile;
float dxc = Math.max(minQuantile, Math.min(maxQuantile, v)) -
minQuantile;
floa
benwtrent commented on code in PR #13910:
URL: https://github.com/apache/lucene/pull/13910#discussion_r1803130185
##
lucene/test-framework/src/java/org/apache/lucene/tests/index/BaseKnnVectorsFormatTestCase.java:
##
@@ -1906,4 +1916,122 @@ public void testMismatchedFields() thro
benwtrent commented on code in PR #13910:
URL: https://github.com/apache/lucene/pull/13910#discussion_r1803132931
##
lucene/test-framework/src/java/org/apache/lucene/tests/index/BaseKnnVectorsFormatTestCase.java:
##
@@ -1906,4 +1916,122 @@ public void testMismatchedFields() thro
benwtrent commented on code in PR #13910:
URL: https://github.com/apache/lucene/pull/13910#discussion_r1803081052
##
lucene/test-framework/src/java/org/apache/lucene/tests/index/BaseKnnVectorsFormatTestCase.java:
##
@@ -1906,4 +1916,122 @@ public void testMismatchedFields() thro
risdenk commented on PR #13827:
URL: https://github.com/apache/lucene/pull/13827#issuecomment-2417258152
Thanks appreciate the second set of eyes. Glad to know I'm not going crazy.
I also didn't see a good way forward for updating the array.
--
This is an automated message from the Apache
risdenk closed pull request #13827: Fixed bit set vector
URL: https://github.com/apache/lucene/pull/13827
--
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
benwtrent commented on PR #13651:
URL: https://github.com/apache/lucene/pull/13651#issuecomment-2417284114
I am currently working on moving this to Lucene101 format with the bug fixes
we discovered in additional testing.
--
This is an automated message from the Apache Git Service.
To res
msokolov commented on code in PR #13910:
URL: https://github.com/apache/lucene/pull/13910#discussion_r1803114051
##
lucene/test-framework/src/java/org/apache/lucene/tests/index/BaseKnnVectorsFormatTestCase.java:
##
@@ -1906,4 +1916,122 @@ public void testMismatchedFields() throw
benwtrent commented on issue #13921:
URL: https://github.com/apache/lucene/issues/13921#issuecomment-2416554406
@hanqiushi you don't need to do anything special. Since its a format, you
only have to query the index.
So, once you have indexed a bunch of
[KnnFloatVectorField](https://
benwtrent closed issue #13921: How to use hnsw int4 when loading index
URL: https://github.com/apache/lucene/issues/13921
--
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 uns
benwtrent commented on PR #13827:
URL: https://github.com/apache/lucene/pull/13827#issuecomment-2416558059
Yeah, I poked around and it just seems like unloading the CPU vectors back
into an array eats all performance gains.
I tried some reorganization to see if there was a better way,
dnhatn closed pull request #13803: Copy stored fields during flush with index
sort
URL: https://github.com/apache/lucene/pull/13803
--
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 comme
dnhatn commented on PR #13803:
URL: https://github.com/apache/lucene/pull/13803#issuecomment-2418198002
I ran the benchmarks multiple times, but the improvement seems to be more
noise than a real gain, so I'm closing this PR.
--
This is an automated message from the Apache Git Service.
To
harshavamsi commented on issue #12686:
URL: https://github.com/apache/lucene/issues/12686#issuecomment-2417446490
Thanks @prudhvigodithi!
--
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 speci
harshavamsi closed issue #12686: `IndexOrDocValuesQuery` does not support query
highlighting
URL: https://github.com/apache/lucene/issues/12686
--
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
mdmarshmallow commented on issue #13913:
URL: https://github.com/apache/lucene/issues/13913#issuecomment-2418316155
After some discussion with @mikemccand, I'm currently planning on making a
separate RAM manager that will force that largest `IndexWriters` to flush their
buffers if the total
27 matches
Mail list logo
