You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@beam.apache.org by tv...@apache.org on 2025/06/11 17:42:13 UTC

(beam) branch master updated: Update CHANGES.md (#35242)

This is an automated email from the ASF dual-hosted git repository.

tvalentyn pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/beam.git


The following commit(s) were added to refs/heads/master by this push:
     new c9958d28114 Update CHANGES.md (#35242)
c9958d28114 is described below

commit c9958d28114a0c65b2fd1086b42039356ecb5aea
Author: tvalentyn <tv...@users.noreply.github.com>
AuthorDate: Wed Jun 11 10:42:06 2025 -0700

    Update CHANGES.md (#35242)
    
    Highlight improvements to the vllm model handler.
---
 CHANGES.md | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)

diff --git a/CHANGES.md b/CHANGES.md
index cc8480c1a7e..fc24e248439 100644
--- a/CHANGES.md
+++ b/CHANGES.md
@@ -69,8 +69,8 @@
 
 ## Highlights
 
-* New highly anticipated feature X added to Python SDK ([#X](https://github.com/apache/beam/issues/X)).
-* New highly anticipated feature Y added to Java SDK ([#Y](https://github.com/apache/beam/issues/Y)).
+* [Python] Several quality-of-life improvements to the vLLM model handler. If you use Beam RunInference with vLLM model handlers, we strongly recommend updating past this release.
+
 
 ## I/Os
 
@@ -101,8 +101,9 @@
 
 * Fixed X (Java/Python) ([#X](https://github.com/apache/beam/issues/X)).
 * (Java) Fixed cassandraIO ReadAll does not let a pipeline handle or retry exceptions ([#34191](https://github.com/apache/beam/pull/34191)).
-* [Python] Fixed vLLM breaks dataflow logging. ([35053](https://github.com/apache/beam/pull/35053))
-* [Python] Fixed vLLM leaks connections causing a throughput bottleneck and underutilization of GPU ([35053](https://github.com/apache/beam/pull/35053))
+* [Python] Fixed vLLM model handlers breaking Beam logging. ([#35053](https://github.com/apache/beam/pull/35053)).
+* [Python] Fixed vLLM connection leaks that caused a throughput bottleneck and underutilization of GPU ([#35053](https://github.com/apache/beam/pull/35053)).
+* [Python] Fixed vLLM server recovery mechanism in the event of a process termination ([#35234](https://github.com/apache/beam/pull/35234)).
 * (Python) Fixed cloudpickle overwriting class states every time loading a same object of dynamic class ([#35062](https://github.com/apache/beam/issues/35062)).
 * [Python] Fixed pip install apache-beam[interactive] causes crash on google colab ([#35148](https://github.com/apache/beam/pull/35148)).