biondizzle/vllm
1
0
Fork 0
Code Issues Pull Requests Actions 2 Packages Projects Releases Wiki Activity

[Hardware][IBM Z]Enable v1 for s390x and s390x dockerfile fixes (#22725)

Browse Source 
Signed-off-by: Nikhil Suryawanshi <suryawanshin74@gmail.com>
This commit is contained in:
Nikhil Suryawanshi
2025-08-19 10:10:37 +05:30
committed by GitHub
parent e9d6a3db69
commit 78dba404ad
7 changed files with 96 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
vllm/engine/arg_utils.py
View File

@@ -1076,12 +1076,13 @@ class EngineArgs:
# Set default arguments for V0 or V1 Engine.
if use_v1:
self._set_default_args_v1(usage_context, model_config)
# Disable chunked prefill for POWER (ppc64le)/ARM CPUs in V1
# Disable chunked prefill for POWER (ppc64le)/ARM/s390x CPUs in V1
if current_platform.is_cpu(
) and current_platform.get_cpu_architecture() in (
CpuArchEnum.POWERPC, CpuArchEnum.ARM):
CpuArchEnum.POWERPC, CpuArchEnum.S390X, CpuArchEnum.ARM):
logger.info(
"Chunked prefill is not supported for ARM and POWER CPUs; "
"Chunked prefill is not supported for ARM and POWER "
"and S390X CPUs; "
"disabling it for V1 backend.")
self.enable_chunked_prefill = False
else: