@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / 409475a827
Fork 0

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 16, 04:30 UTC

[Bugfix] Fix issues in CPU build Dockerfile. Fixes #9182 (#11435)

Signed-off-by: Yuan Tang <terrytangyuan@gmail.com>

Yuan Tang committed Dec 24, 2024 at 11:53 UTC 409475a827795000301f0d27582befab0563888b
2 files changed +10 -6
Dockerfile.cpu
+3 -3
@@ -26,10 +26,10 @@ RUN pip install intel_extension_for_pytorch==2.5.0
26
27 WORKDIR /workspace
28
29 +COPY requirements-build.txt requirements-build.txt
30 ARG PIP_EXTRA_INDEX_URL="https://download.pytorch.org/whl/cpu"
31 ENV PIP_EXTRA_INDEX_URL=${PIP_EXTRA_INDEX_URL}
32 RUN --mount=type=cache,target=/root/.cache/pip \
32 - --mount=type=bind,src=requirements-build.txt,target=requirements-build.txt \
33 pip install --upgrade pip && \
34 pip install -r requirements-build.txt
35
@@ -37,9 +37,9 @@ FROM cpu-test-1 AS build
37
38 WORKDIR /workspace/vllm
39
40 +COPY requirements-common.txt requirements-common.txt
41 +COPY requirements-cpu.txt requirements-cpu.txt
42 RUN --mount=type=cache,target=/root/.cache/pip \
41 - --mount=type=bind,src=requirements-common.txt,target=requirements-common.txt \
42 - --mount=type=bind,src=requirements-cpu.txt,target=requirements-cpu.txt \
43 pip install -v -r requirements-cpu.txt
44
45 COPY . .
setup.py
+7 -3
@@ -455,9 +455,13 @@ def get_gaudi_sw_version():
455
456
457 def get_vllm_version() -> str:
458 - version = get_version(
459 - write_to="vllm/_version.py", # TODO: move this to pyproject.toml
460 - )
458 + # TODO: Revisit this temporary approach: https://github.com/vllm-project/vllm/issues/9182#issuecomment-2404860236
459 + try:
460 + version = get_version(
461 + write_to="vllm/_version.py", # TODO: move this to pyproject.toml
462 + )
463 + except LookupError:
464 + version = "0.0.0"
465
466 sep = "+" if "+" not in version else "." # dev versions might contain +
467