A high-throughput and memory-efficient inference and serving engine for LLMs
Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git
Synced Nov 14, 05:30 UTC
[fix] Update torch version in cpu-build.txt for AArch64/ppc64le and Darwin (#25579)
Signed-off-by: Fadi Arafeh <fadi.arafeh@arm.com>
Fadi Arafeh committed
Sep 25, 2025 at 10:39 UTC
393de22d2ebf0ba397df5ef0bb9b7b9111a3d190
1 file changed
+2
-1
requirements/cpu-build.txt
+2
-1
@@ -6,7 +6,8 @@ packaging>=24.2
6
setuptools>=77.0.3,<80.0.0
7
setuptools-scm>=8
8
--extra-index-url https://download.pytorch.org/whl/cpu
9
-torch==2.6.0+cpu
9
+torch==2.6.0+cpu; platform_machine == "x86_64" # torch>2.6.0+cpu has performance regression on x86 platform, see https://github.com/pytorch/pytorch/pull/151218
10
+torch==2.8.0; platform_machine == "ppc64le" or platform_machine == "aarch64" or platform_system == "Darwin"
11
wheel
12
jinja2>=3.1.6
13
regex