Script 'mail_helper' called by obssrc
Hello community,

here is the log from the commit of package python-vllm for openSUSE:Factory 
checked in at 2026-07-26 11:30:13
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Comparing /work/SRC/openSUSE:Factory/python-vllm (Old)
 and      /work/SRC/openSUSE:Factory/.python-vllm.new.2004 (New)
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Package is "python-vllm"

Sun Jul 26 11:30:13 2026 rev:4 rq:1367774 version:0.26.0

Changes:
--------
--- /work/SRC/openSUSE:Factory/python-vllm/python-vllm.changes  2026-07-22 
19:08:49.556704332 +0200
+++ /work/SRC/openSUSE:Factory/.python-vllm.new.2004/python-vllm.changes        
2026-07-26 11:32:48.203661604 +0200
@@ -1,0 +2,24 @@
+Sat Jul 25 19:30:45 UTC 2026 - Martin Pluskal <[email protected]>
+
+- Update to version 0.26.0:
+  * New Inkling model family support including speculative
+    decoding, LoRA and quantization; further new models:
+    BertForMaskedLM, Roberta/XLMRoberta token classification,
+    Cosmos3 Edge Reasoner, TranslateGemma-12b-it
+  * DeepSeek-V4 performance push: specialized routing kernel,
+    fused_topk_bias kernel, sparse decode/prefill optimizations
+  * fp32 lm_head for generation models via head_dtype, also on
+    the LoRA path
+  * Attention backend selectable per KV-cache group and
+    sliding-window support as an explicit backend capability
+  * Matured KV offloading and tiered secondary storage:
+    offloading metrics, object-store secondary tier,
+    encoder-cache connectors with CPU offloading
+  * Transformers 5.13 support with more models on the
+    Transformers modeling backend
+  * Requirements: diskcache and flashinfer-cubin dropped
+- Rebase vllm-relax-cpu-requirements.patch onto the 0.26.0
+  requirements (diskcache line removed upstream)
+- Drop the python-diskcache runtime requirement accordingly
+
+-------------------------------------------------------------------

Old:
----
  vllm-0.25.1.tar.gz

New:
----
  vllm-0.26.0.tar.gz

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Other differences:
------------------
++++++ python-vllm.spec ++++++
--- /var/tmp/diff_new_pack.esEi4U/_old  2026-07-26 11:32:52.411806589 +0200
+++ /var/tmp/diff_new_pack.esEi4U/_new  2026-07-26 11:32:52.415806727 +0200
@@ -29,7 +29,7 @@
 # compiled kernels is tracked as a follow-up once that is resolved.
 %define vllm_target_device empty
 Name:           python-vllm
-Version:        0.25.1
+Version:        0.26.0
 Release:        0
 Summary:        A high-throughput and memory-efficient inference and serving 
engine for LLMs
 License:        Apache-2.0
@@ -60,6 +60,8 @@
 # import torch dlopens libopenblas.so.0 at build time.
 BuildRequires:  libopenblas_pthreads0
 BuildRequires:  python-rpm-macros
+Requires:       python-Pillow
+Requires:       python-PyYAML
 Requires:       python-aiohttp >= 3.13.3
 Requires:       python-anthropic >= 0.71.0
 Requires:       python-blake3
@@ -68,7 +70,6 @@
 Requires:       python-cloudpickle
 Requires:       python-compressed-tensors = 0.17.1
 Requires:       python-depyf = 0.20.0
-Requires:       python-diskcache = 5.6.3
 Requires:       python-einops
 Requires:       python-fastapi >= 0.133.0
 Requires:       python-filelock >= 3.16.1
@@ -93,9 +94,8 @@
 Requires:       python-opentelemetry-semantic-conventions-ai >= 0.4.1
 Requires:       python-outlines_core = 0.2.14
 Requires:       python-partial-json-parser
-Requires:       python-Pillow
-Requires:       python-prometheus-fastapi-instrumentator >= 8.0.0
 Requires:       python-prometheus-client >= 0.18.0
+Requires:       python-prometheus-fastapi-instrumentator >= 8.0.0
 Requires:       python-protobuf >= 5.29.6
 Requires:       python-psutil
 Requires:       python-py-cpuinfo
@@ -103,7 +103,6 @@
 Requires:       python-pydantic >= 2.12.0
 Requires:       python-pydantic-extra-types
 Requires:       python-python-json-logger
-Requires:       python-PyYAML
 Requires:       python-pyzmq >= 25.0.0
 Requires:       python-regex
 Requires:       python-requests >= 2.26.0

++++++ vllm-0.25.1.tar.gz -> vllm-0.26.0.tar.gz ++++++
/work/SRC/openSUSE:Factory/python-vllm/vllm-0.25.1.tar.gz 
/work/SRC/openSUSE:Factory/.python-vllm.new.2004/vllm-0.26.0.tar.gz differ: 
char 5, line 1

++++++ vllm-relax-cpu-requirements.patch ++++++
--- /var/tmp/diff_new_pack.esEi4U/_old  2026-07-26 11:32:52.467808518 +0200
+++ /var/tmp/diff_new_pack.esEi4U/_new  2026-07-26 11:32:52.475808794 +0200
@@ -1,5 +1,5 @@
---- a/requirements/cpu.txt     2026-07-13 08:06:59.821793879 +0200
-+++ b/requirements/cpu.txt     2026-07-13 08:07:15.327541418 +0200
+--- a/requirements/cpu.txt
++++ b/requirements/cpu.txt
 @@ -1,25 +1,13 @@
  # Common dependencies
  -r common.txt
@@ -30,12 +30,12 @@
  
  # Use this to gather CPU info and optimize based on ARM Neoverse cores
  py-cpuinfo; platform_machine == "aarch64"
---- a/requirements/common.txt  2026-07-13 08:06:59.821793879 +0200
-+++ b/requirements/common.txt  2026-07-13 08:07:24.948625739 +0200
-@@ -25,7 +25,7 @@
+--- a/requirements/common.txt
++++ b/requirements/common.txt
+@@ -23,7 +23,7 @@
+ lm-format-enforcer == 0.11.3
+ llguidance >= 1.7.0, < 1.8.0; platform_machine == "x86_64" or 
platform_machine == "arm64" or platform_machine == "aarch64" or 
platform_machine == "ppc64le" or platform_machine == "s390x"
  outlines_core == 0.2.14
- # required for outlines backend disk cache
- diskcache == 5.6.3
 -lark == 1.2.2
 +lark >= 1.2.2
  xgrammar >= 0.2.1, < 1.0.0; platform_machine == "x86_64" or platform_machine 
== "aarch64" or platform_machine == "arm64" or platform_machine == "s390x" or 
platform_machine == "ppc64le"

Reply via email to