benwtrent commented on issue #14342:
URL: https://github.com/apache/lucene/issues/14342#issuecomment-2733531171
First, thank you @lpld for digging in and running these benchmarks!
OK, I think I see the weirdness with the `mnist` data set. Its not about it
being a transformer model, it
javanna commented on code in PR #14364:
URL: https://github.com/apache/lucene/pull/14364#discussion_r2000898675
##
lucene/suggest/src/test/org/apache/lucene/search/suggest/document/TestSuggestField.java:
##
@@ -951,7 +951,16 @@ static IndexWriterConfig iwcWithSuggestField(Analyz
gf2121 commented on PR #14365:
URL: https://github.com/apache/lucene/pull/14365#issuecomment-2733285724
I'm seeing even results on `wikimediumall`
```
TaskQPS baseline StdDevQPS
my_modified_version StdDevPct diff p-value
dweiss commented on code in PR #14350:
URL: https://github.com/apache/lucene/pull/14350#discussion_r2000465944
##
lucene/core/src/java/org/apache/lucene/util/automaton/Automata.java:
##
@@ -608,7 +608,24 @@ public static Automaton makeStringUnion(Iterable
utf8Strings) {
if
gf2121 closed pull request #13521: Introduce new encoding of BPV 21 for
DocIdsWriter used in BKD Tree
URL: https://github.com/apache/lucene/pull/13521
--
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
gf2121 merged PR #14361:
URL: https://github.com/apache/lucene/pull/14361
--
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.apac
gf2121 commented on PR #13521:
URL: https://github.com/apache/lucene/pull/13521#issuecomment-2732037641
Closing this in favor of https://github.com/apache/lucene/pull/14361.
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and
vigyasharma commented on issue #14362:
URL: https://github.com/apache/lucene/issues/14362#issuecomment-2731919938
Hi @guojialiang92, Could you elaborate more on how you plan to use this
capability? It's not immediately obvious why modifying `segmentInfos.counter`
will help with peer recover
rmuir commented on code in PR #14350:
URL: https://github.com/apache/lucene/pull/14350#discussion_r2002272398
##
lucene/core/src/java/org/apache/lucene/util/automaton/CaseFolding.java:
##
@@ -743,4 +743,42 @@ static int[] lookupAlternates(int codepoint) {
return alts;
vigyasharma commented on code in PR #14335:
URL: https://github.com/apache/lucene/pull/14335#discussion_r2000345662
##
lucene/core/src/test/org/apache/lucene/index/TestMultiTenantMergeScheduler.java:
##
@@ -0,0 +1,73 @@
+package org.apache.lucene.index;
+
+import org.apache.luce
stefanvodita commented on issue #13898:
URL: https://github.com/apache/lucene/issues/13898#issuecomment-2733970610
Just to clarify - the restriction @dweiss mentioned applies to the
`changelog-enforcer` action, but not to the `checkout` action we are using.
@pseudo-nymous - I'm seeing
jpountz commented on PR #14365:
URL: https://github.com/apache/lucene/pull/14365#issuecomment-2734597577
Maybe we should stop only adding doc IDs to the `BulkAdder` if they are
greater than the max collected doc so far. Skipping these doc IDs looks like it
hurts vectorization, I played with
msfroh commented on code in PR #14350:
URL: https://github.com/apache/lucene/pull/14350#discussion_r2002184400
##
lucene/core/src/java/org/apache/lucene/util/automaton/CaseFolding.java:
##
@@ -743,4 +743,42 @@ static int[] lookupAlternates(int codepoint) {
return alts;
benwtrent commented on issue #14342:
URL: https://github.com/apache/lucene/issues/14342#issuecomment-2734567137
OK, a colleague and I spent some time digging into this and Option 0 (a bug)
turned out to be the case. Its a 5 character change (like all good bugs), but
here are the new recall
pseudo-nymous opened a new pull request, #14369:
URL: https://github.com/apache/lucene/pull/14369
### Description
This pull request contains a fix for changelog automation that has been
added recently. We have seen failures where either diff calculation wrt base
commit was wrong or base
pseudo-nymous commented on issue #13898:
URL: https://github.com/apache/lucene/issues/13898#issuecomment-2735279420
@stefanvodita I have added a fix for it. Please take a look.
https://github.com/apache/lucene/pull/14369
--
This is an automated message from the Apache Git Service.
To res
pseudo-nymous commented on PR #14369:
URL: https://github.com/apache/lucene/pull/14369#issuecomment-2735285383
We can fetch all history using checkout actions itself using flag
`fetch-depth: 0`. But it fetches all the history for all branches and tags
which is not required here.
--
This
gf2121 commented on PR #14365:
URL: https://github.com/apache/lucene/pull/14365#issuecomment-2735314110
Thanks for running benchmark, the speed up is great!
> Skipping these doc IDs looks like it hurts vectorization, I played with
disabling these if statements locally and get a good s
amosbird opened a new pull request, #14368:
URL: https://github.com/apache/lucene/pull/14368
### Description
Update prefetch heuristic comments to reflect that skip data is now inlined
into postings lists.
--
This is an automated message from the Apache Git Service.
To
jpountz merged PR #14368:
URL: https://github.com/apache/lucene/pull/14368
--
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
vigyasharma commented on code in PR #14335:
URL: https://github.com/apache/lucene/pull/14335#discussion_r2001826532
##
lucene/core/src/java/org/apache/lucene/index/MultiTenantMergeScheduler.java:
##
@@ -0,0 +1,70 @@
+package org.apache.lucene.index;
+
+import java.util.concurren
jpountz commented on PR #14368:
URL: https://github.com/apache/lucene/pull/14368#issuecomment-2734638261
Good catch!
--
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 unsu
22 matches
Mail list logo
