jpountz commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r783701570
##
File path: lucene/core/src/java/org/apache/lucene/codecs/KnnVectorsWriter.java
##
@@ -49,7 +43,7 @@ public abstract void writeField(FieldInfo fieldInfo,
jpountz commented on pull request #592:
URL: https://github.com/apache/lucene/pull/592#issuecomment-1011911664
Interesting, so it looks like rolling the loop back actually helps a bit.
Can you open a new JIRA/PR for this change specifically @gf2121 ?
--
This is an automated message from
Feng Guo created LUCENE-10376:
-
Summary: Roll up the loop in vint/vlong in DataInput
Key: LUCENE-10376
URL: https://issues.apache.org/jira/browse/LUCENE-10376
Project: Lucene - Core
Issue Type: I
jpountz commented on a change in pull request #541:
URL: https://github.com/apache/lucene/pull/541#discussion_r783744532
##
File path: lucene/core/src/java/org/apache/lucene/util/bkd/BKDForUtil.java
##
@@ -0,0 +1,88 @@
+// This file has been automatically generated, DO NOT EDIT
[
https://issues.apache.org/jira/browse/LUCENE-10376?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475181#comment-17475181
]
Adrien Grand commented on LUCENE-10376:
---
+1
> Roll up the loop in vint/vlong in
gf2121 opened a new pull request #602:
URL: https://github.com/apache/lucene/pull/602
https://issues.apache.org/jira/browse/LUCENE-10376
--
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
gf2121 commented on pull request #592:
URL: https://github.com/apache/lucene/pull/592#issuecomment-1011957978
Thanks @jpountz . I opened
https://issues.apache.org/jira/browse/LUCENE-10376
(https://github.com/apache/lucene/pull/602)
It can be seen that `IntNRQ` is getting a 10% speed
gf2121 edited a comment on pull request #592:
URL: https://github.com/apache/lucene/pull/592#issuecomment-1011957978
Thanks @jpountz . I opened
https://issues.apache.org/jira/browse/LUCENE-10376
(https://github.com/apache/lucene/pull/602)
It can be seen that `IntNRQ` is getting a 10
rmuir commented on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1011980920
There are some more copies of this stuff in BufferedIndexInput and
BlockPackedReaderIterator
--
This is an automated message from the Apache Git Service.
To respond to the message
iverase commented on pull request #541:
URL: https://github.com/apache/lucene/pull/541#issuecomment-1012044387
I don't like that the optimisation only works for a specific number of
points, it feels very tricky. If we are doing something like that we should
remove the possibility of users
iverase edited a comment on pull request #541:
URL: https://github.com/apache/lucene/pull/541#issuecomment-1012044387
I don't like that the optimisation only works for a specific number of
points, it feels trappy. If we are doing something like that we should remove
the possibility of user
Alan Woodward created LUCENE-10377:
--
Summary: Replace 'sortPos' parameter in SortField.getComparator()
Key: LUCENE-10377
URL: https://issues.apache.org/jira/browse/LUCENE-10377
Project: Lucene - Core
Adrien Grand created LUCENE-10378:
-
Summary: Implement Weight#count on PointRangeQuery
Key: LUCENE-10378
URL: https://issues.apache.org/jira/browse/LUCENE-10378
Project: Lucene - Core
Issue T
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: (was: LUCENE-10367.patch)
> Use WANDScorer in CoveringQuery Can accelerate scorer
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: LUCENE-10367.patch
> Use WANDScorer in CoveringQuery Can accelerate scorer time
> ---
romseygeek commented on pull request #603:
URL: https://github.com/apache/lucene/pull/603#issuecomment-1012140425
This might make it possible to remove the `getPointsValues()` override from
NumericLeafComparator as well as a follow-up, cc @mayya-sharipova
--
This is an automated message
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: TestCoveringQueryBench.java
> Use WANDScorer in CoveringQuery Can accelerate scorer t
jpountz commented on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1012141279
There is ByteArrayDataInput too.
--
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 th
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475364#comment-17475364
]
Adrien Grand commented on LUCENE-10367:
---
Your patch is only rewriting to a Boolea
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475365#comment-17475365
]
LuYunCheng commented on LUCENE-10367:
-
[~jpountz] Yes, It's better to handled this
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475370#comment-17475370
]
LuYunCheng commented on LUCENE-10367:
-
> Your patch is only rewriting to a Boolean
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: LUCENE-10367-1.patch
> Use WANDScorer in CoveringQuery Can accelerate scorer time
> -
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475379#comment-17475379
]
LuYunCheng commented on LUCENE-10367:
-
i think when the minimumNumberMatch is Const
[
https://issues.apache.org/jira/browse/LUCENE-10374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475381#comment-17475381
]
Robert Muir commented on LUCENE-10374:
--
Looks like that did the trick!
http://peo
gf2121 commented on pull request #541:
URL: https://github.com/apache/lucene/pull/541#issuecomment-1012163438
Thanks @iverase !
> I remember in the approach I tried I was batching the docIds by a number
(128 or 256) but in general tI didn't see much better performance comparing to
t
mayya-sharipova commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784021962
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(F
mayya-sharipova commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784021962
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(F
mayya-sharipova commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784021962
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(F
mayya-sharipova commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784021962
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(F
mayya-sharipova commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784037293
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(F
[
https://issues.apache.org/jira/browse/LUCENE-10288?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475423#comment-17475423
]
Adrien Grand commented on LUCENE-10288:
---
This problem would also happen if you us
gsmiller commented on pull request #597:
URL: https://github.com/apache/lucene/pull/597#issuecomment-1012217406
Interesting. It appears LUCENE-10372 was the cause of the regression in the
nightly benchmarks (see most recent runs of the tasks mentioned
[above](https://github.com/apache/luce
dweiss commented on pull request #597:
URL: https://github.com/apache/lucene/pull/597#issuecomment-1012219388
> LUCENE-10372 when run on our machines while the nightly run would show a
regression.
I don't think it's a tractable problem in general, given hardware
architecture and rac
gsmiller edited a comment on pull request #597:
URL: https://github.com/apache/lucene/pull/597#issuecomment-1012217406
Interesting. It appears LUCENE-10372 was the cause of the regression in the
nightly benchmarks (see most recent runs of the tasks mentioned
[above](https://github.com/apac
[
https://issues.apache.org/jira/browse/LUCENE-10374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475432#comment-17475432
]
Greg Miller commented on LUCENE-10374:
--
Indeed [~rcmuir]. Now I'm even more confus
mayya-sharipova commented on pull request #601:
URL: https://github.com/apache/lucene/pull/601#issuecomment-1012221563
@jtibshirani That's a great change. Agree with Adrien, great to see these
speedups with an additional benefit of simplified code.
--
This is an automated message from th
gsmiller edited a comment on pull request #597:
URL: https://github.com/apache/lucene/pull/597#issuecomment-1012217406
Interesting. It appears LUCENE-10350 was the cause of the regression in the
nightly benchmarks (see most recent runs of the tasks mentioned
[above](https://github.com/apac
[
https://issues.apache.org/jira/browse/LUCENE-10374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475435#comment-17475435
]
Robert Muir commented on LUCENE-10374:
--
[~gsmiller] I agree, looking at the commit
[
https://issues.apache.org/jira/browse/LUCENE-10374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475438#comment-17475438
]
Greg Miller commented on LUCENE-10374:
--
It appears that LUCENE-10350 was responsib
[
https://issues.apache.org/jira/browse/LUCENE-10375?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475446#comment-17475446
]
Michael Sokolov commented on LUCENE-10375:
--
Ooh, exciting. That code was compl
msokolov commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784071103
##
File path: lucene/core/src/java/org/apache/lucene/codecs/KnnVectorsWriter.java
##
@@ -271,52 +207,5 @@ public long cost() {
public int dimension() {
gsmiller opened a new pull request #604:
URL: https://github.com/apache/lucene/pull/604
This reverts LUCENE-10350 on 9.x as well given that it appears to be the
cause of a performance regression (see #597)
--
This is an automated message from the Apache Git Service.
To respond to the mes
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475507#comment-17475507
]
Adrien Grand commented on LUCENE-10367:
---
Thanks for running benchmarks. It looks
[
https://issues.apache.org/jira/browse/LUCENE-10374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475529#comment-17475529
]
Greg Miller commented on LUCENE-10374:
--
OK, here's a thought: LUCENE-10350 actuall
gf2121 commented on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1012314786
Thanks f! I've fixed these places and will do a benchmark based on the
newest code tonight.
--
This is an automated message from the Apache Git Service.
To respond to the message
gf2121 edited a comment on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1012314786
Thanks ! I've fixed these places and will do a benchmark based on the newest
code tonight.
--
This is an automated message from the Apache Git Service.
To respond to the m
Greg Miller created LUCENE-10379:
Summary: Count directly into the values array in
FastTaxonomyFacetCounts#countAl
Key: LUCENE-10379
URL: https://issues.apache.org/jira/browse/LUCENE-10379
Project: Lu
gsmiller opened a new pull request #605:
URL: https://github.com/apache/lucene/pull/605
This change just pulls out part of the optimization done (and subsequently
reverted) in LUCENE-10350. The idea is to push only part of the optimization to
see if we can isolate the performance regressio
gsmiller merged pull request #604:
URL: https://github.com/apache/lucene/pull/604
--
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.
[
https://issues.apache.org/jira/browse/LUCENE-10350?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475544#comment-17475544
]
ASF subversion and git services commented on LUCENE-10350:
--
Co
[
https://issues.apache.org/jira/browse/LUCENE-10372?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475545#comment-17475545
]
Greg Miller commented on LUCENE-10372:
--
This is now reverted on both {{main}} and
[
https://issues.apache.org/jira/browse/LUCENE-10372?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Greg Miller resolved LUCENE-10372.
--
Fix Version/s: 9.1
10.0 (main)
Resolution: Fixed
> Performance of T
uschindler commented on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1012341843
Great that the original code was still there in a comment. 🤗 This makes easy
to review.
I don't fully understand the variant that supports negative values. Do we
still n
gsmiller commented on pull request #605:
URL: https://github.com/apache/lucene/pull/605#issuecomment-1012342391
I'm seeing the nice performance improvement locally that @gf2121 and I both
observed with LUCENE-10350, so I'm going to go ahead and merge this, which is
just a portion of LUCENE
gsmiller merged pull request #605:
URL: https://github.com/apache/lucene/pull/605
--
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.
[
https://issues.apache.org/jira/browse/LUCENE-10379?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475547#comment-17475547
]
ASF subversion and git services commented on LUCENE-10379:
--
Co
[
https://issues.apache.org/jira/browse/LUCENE-10374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475548#comment-17475548
]
Greg Miller commented on LUCENE-10374:
--
Just merged a change for just part 1 above
[
https://issues.apache.org/jira/browse/LUCENE-10379?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475550#comment-17475550
]
Greg Miller commented on LUCENE-10379:
--
Merged this and will watch the next nightl
[
https://issues.apache.org/jira/browse/LUCENE-10374?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475553#comment-17475553
]
Greg Miller commented on LUCENE-10374:
--
{quote}[~gsmiller] I agree, looking at the
gf2121 commented on pull request #605:
URL: https://github.com/apache/lucene/pull/605#issuecomment-1012355358
This is great! Thanks @gsmiller !
--
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 th
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784189861
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(Field
gf2121 commented on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1012371355
Thanks @uschindler .
> I don't fully understand the variant that supports negative values. Do we
still need it as the default one again supports negative values. The code loo
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784199491
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(Field
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784199665
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(Field
[
https://issues.apache.org/jira/browse/LUCENE-10374?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Greg Miller reassigned LUCENE-10374:
Assignee: Greg Miller
> Track down the "browse" taxonomy faceting qps regression
> -
[
https://issues.apache.org/jira/browse/LUCENE-10379?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Greg Miller reassigned LUCENE-10379:
Assignee: Greg Miller
> Count directly into the values array in FastTaxonomyFacetCounts#
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784214402
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -155,6 +207,37 @@ public void writeField(Field
gerlowskija commented on pull request #2634:
URL: https://github.com/apache/lucene-solr/pull/2634#issuecomment-1012399441
I ended up targeting 9.0 with these changes, as it _looks_ like 8.x is about
wrapped up except for urgent security fixes.
With that in mind I'm going to close thi
gerlowskija closed pull request #2634:
URL: https://github.com/apache/lucene-solr/pull/2634
--
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
[
https://issues.apache.org/jira/browse/LUCENE-9952?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
Greg Miller reassigned LUCENE-9952:
---
Assignee: Greg Miller
> FacetResult#value can be inaccurate in SortedSetDocValueFacetCounts
[
https://issues.apache.org/jira/browse/LUCENE-9952?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475634#comment-17475634
]
Greg Miller commented on LUCENE-9952:
-
I think we're finally setup to address this n
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784251362
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(Field
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784270583
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(Field
[
https://issues.apache.org/jira/browse/LUCENE-9952?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475682#comment-17475682
]
Greg Miller commented on LUCENE-9952:
-
OK, I've got a fix for this working but am go
mayya-sharipova commented on a change in pull request #416:
URL: https://github.com/apache/lucene/pull/416#discussion_r784284382
##
File path: lucene/core/src/java/org/apache/lucene/util/hnsw/HnswGraph.java
##
@@ -56,75 +57,124 @@
public final class HnswGraph extends KnnGraphV
mayya-sharipova commented on a change in pull request #416:
URL: https://github.com/apache/lucene/pull/416#discussion_r784284582
##
File path: lucene/core/src/java/org/apache/lucene/util/hnsw/HnswGraph.java
##
@@ -56,75 +57,124 @@
public final class HnswGraph extends KnnGraphV
mayya-sharipova commented on a change in pull request #416:
URL: https://github.com/apache/lucene/pull/416#discussion_r784284757
##
File path: lucene/core/src/java/org/apache/lucene/util/hnsw/HnswGraph.java
##
@@ -40,10 +41,10 @@
* Hyperparameters
*
*
- * numSeed is th
mayya-sharipova commented on a change in pull request #416:
URL: https://github.com/apache/lucene/pull/416#discussion_r784284891
##
File path: lucene/core/src/test/org/apache/lucene/util/hnsw/TestHnswGraph.java
##
@@ -116,33 +116,59 @@ public KnnVectorsFormat
getKnnVectorsForm
mayya-sharipova commented on a change in pull request #416:
URL: https://github.com/apache/lucene/pull/416#discussion_r784285235
##
File path: lucene/core/src/test/org/apache/lucene/index/TestKnnGraph.java
##
@@ -153,21 +161,56 @@ public void testMergeProducesSameGraph() throws
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784334763
##
File path: lucene/core/src/java/org/apache/lucene/codecs/KnnVectorsWriter.java
##
@@ -49,7 +43,7 @@ public abstract void writeField(FieldInfo fieldInf
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784338686
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -145,6 +139,64 @@ public void writeField(Field
jtibshirani commented on pull request #601:
URL: https://github.com/apache/lucene/pull/601#issuecomment-1012533036
Thanks for the reviews. For each comment I either pushed a commit or asked a
follow-up question.
--
This is an automated message from the Apache Git Service.
To respond to t
uschindler commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784344280
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -155,6 +207,37 @@ public void writeField(FieldI
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784346369
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -138,32 +129,145 @@ public void writeField(Fie
jtibshirani commented on pull request #588:
URL: https://github.com/apache/lucene/pull/588#issuecomment-1012582171
> Maybe the best strategy here would be to keep the change entry for
whichever version that goes out first (I think it will be 9.0.1? ), and for the
other one we can skip the
[
https://issues.apache.org/jira/browse/LUCENE-10375?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475818#comment-17475818
]
Julie Tibshirani commented on LUCENE-10375:
---
I was happily surprised it helpe
uschindler commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784389033
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -138,32 +129,145 @@ public void writeField(Fiel
mayya-sharipova commented on a change in pull request #603:
URL: https://github.com/apache/lucene/pull/603#discussion_r784378305
##
File path:
lucene/core/src/java/org/apache/lucene/search/SortedNumericSortField.java
##
@@ -242,12 +241,21 @@ public void setMissingValue(Object
mayya-sharipova commented on pull request #603:
URL: https://github.com/apache/lucene/pull/603#issuecomment-1012600488
> This might make it possible to remove the getPointsValues() override from
NumericLeafComparator as well as a follow-up, cc @mayya-sharipova
@romseygeek I think you
jtibshirani commented on a change in pull request #601:
URL: https://github.com/apache/lucene/pull/601#discussion_r784393289
##
File path:
lucene/core/src/java/org/apache/lucene/codecs/lucene90/Lucene90HnswVectorsWriter.java
##
@@ -138,32 +129,145 @@ public void writeField(Fie
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Status: Patch Available (was: Open)
> Use WANDScorer in CoveringQuery Can accelerate scorer time
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: (was: LUCENE-10367-1.patch)
> Use WANDScorer in CoveringQuery Can accelerate scor
gf2121 commented on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1012797333
Here is the report (30 JVM * 200 repeat):
```
TaskQPS baseline StdDevQPS
my_modified_version StdDevPct diff p-value
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: (was: LUCENE-10367.patch)
> Use WANDScorer in CoveringQuery Can accelerate scorer
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: LUCENE-10367.patch
> Use WANDScorer in CoveringQuery Can accelerate scorer time
> ---
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: (was: LUCENE-10367.patch)
> Use WANDScorer in CoveringQuery Can accelerate scorer
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
]
LuYunCheng updated LUCENE-10367:
Attachment: LUCENE-10367.patch
> Use WANDScorer in CoveringQuery Can accelerate scorer time
> ---
[
https://issues.apache.org/jira/browse/LUCENE-10367?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=17475977#comment-17475977
]
LuYunCheng commented on LUCENE-10367:
-
[~jpountz] Thanks for reviewing this patch.
gf2121 commented on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1012836713
FYI here are some of my thoughts based on all these benchmark reports:
* Rolling up loops for vint/vlong seems not bring a significant speed up
(neither significant regressio
gf2121 edited a comment on pull request #602:
URL: https://github.com/apache/lucene/pull/602#issuecomment-1012836713
FYI here are some of my thoughts based on all these benchmark reports:
* Rolling up loops for vint/vlong seems not bring a significant speed up
(neither significant re
1 - 100 of 104 matches
Mail list logo
