Skip to content
This repository was archived by the owner on Sep 4, 2025. It is now read-only.

Commit 91e4a51

Browse files
authored
Merge pull request #6 from z103cb/ibm_main_docker_ubi_updates
[CI/Build] Dockerfile.ubi : Remove test stage
2 parents 1e44f5e + 769521f commit 91e4a51

File tree

1 file changed

+0
-21
lines changed

1 file changed

+0
-21
lines changed

Dockerfile.ubi

Lines changed: 0 additions & 21 deletions
Original file line numberDiff line numberDiff line change
@@ -219,27 +219,6 @@ WORKDIR /usr/src/flash-attention-v2
219219
RUN pip --verbose wheel flash-attn==${FLASH_ATTN_VERSION} \
220220
--no-build-isolation --no-deps --no-cache-dir
221221

222-
223-
## Test ########################################################################
224-
FROM dev AS test
225-
226-
WORKDIR /vllm-workspace
227-
# ADD is used to preserve directory structure
228-
# NB: Could leak secrets from local context, the test image should not be pushed
229-
# to a registry
230-
ADD . /vllm-workspace/
231-
# copy pytorch extensions separately to avoid having to rebuild
232-
# when python code changes
233-
COPY --from=build /workspace/vllm/*.so /vllm-workspace/vllm/
234-
# Install flash attention (from pre-built wheel)
235-
RUN --mount=type=bind,from=flash-attn-builder,src=/usr/src/flash-attention-v2,target=/usr/src/flash-attention-v2 \
236-
pip install /usr/src/flash-attention-v2/*.whl --no-cache-dir
237-
# ignore build dependencies installation because we are using pre-complied extensions
238-
RUN rm pyproject.toml
239-
RUN --mount=type=cache,target=/root/.cache/pip \
240-
VLLM_USE_PRECOMPILED=1 pip install . --verbose
241-
242-
243222
## Proto Compilation ###########################################################
244223
FROM python-base AS gen-protos
245224

0 commit comments

Comments
 (0)