diff --git a/docker/Dockerfile b/docker/Dockerfile index 71cef521bed..6e80f3b5965 100644 --- a/docker/Dockerfile +++ b/docker/Dockerfile @@ -418,10 +418,6 @@ RUN --mount=type=cache,target=/root/.cache/uv \ fi && \ python3 setup.py bdist_wheel --dist-dir=dist --py-limited-api=cp38 -# Copy extension wheels from extensions-build stage for later use -COPY --from=extensions-build /tmp/deepgemm/dist /tmp/deepgemm/dist -COPY --from=extensions-build /tmp/ep_kernels_workspace/dist /tmp/ep_kernels_workspace/dist - # Check the size of the wheel if RUN_WHEEL_CHECK is true COPY .buildkite/check-wheel-size.py check-wheel-size.py # sync the default value with .buildkite/check-wheel-size.py @@ -660,9 +656,9 @@ RUN --mount=type=cache,target=/root/.cache/uv \ . /etc/environment && \ uv pip list -# Install deepgemm wheel that has been built in the `build` stage +# Install deepgemm wheel that has been built in the `extensions-build` stage RUN --mount=type=cache,target=/root/.cache/uv \ - --mount=type=bind,from=build,source=/tmp/deepgemm/dist,target=/tmp/deepgemm/dist,ro \ + --mount=type=bind,from=extensions-build,source=/tmp/deepgemm/dist,target=/tmp/deepgemm/dist,ro \ sh -c 'if ls /tmp/deepgemm/dist/*.whl >/dev/null 2>&1; then \ uv pip install --system /tmp/deepgemm/dist/*.whl; \ else \ @@ -672,8 +668,8 @@ RUN --mount=type=cache,target=/root/.cache/uv \ # Pytorch now installs NVSHMEM, setting LD_LIBRARY_PATH ENV LD_LIBRARY_PATH=/usr/local/cuda/lib64:$LD_LIBRARY_PATH -# Install EP kernels wheels (pplx-kernels and DeepEP) that have been built in the `build` stage -RUN --mount=type=bind,from=build,src=/tmp/ep_kernels_workspace/dist,target=/vllm-workspace/ep_kernels/dist \ +# Install EP kernels wheels (pplx-kernels and DeepEP) that have been built in the `extensions-build` stage +RUN --mount=type=bind,from=extensions-build,src=/tmp/ep_kernels_workspace/dist,target=/vllm-workspace/ep_kernels/dist \ --mount=type=cache,target=/root/.cache/uv \ uv pip install --system ep_kernels/dist/*.whl --verbose \ --extra-index-url ${PYTORCH_CUDA_INDEX_BASE_URL}/cu$(echo $CUDA_VERSION | cut -d. -f1,2 | tr -d '.') diff --git a/docs/assets/contributing/dockerfile-stages-dependency.png b/docs/assets/contributing/dockerfile-stages-dependency.png index 9ac394d4c9f..156f0d9b9b7 100644 Binary files a/docs/assets/contributing/dockerfile-stages-dependency.png and b/docs/assets/contributing/dockerfile-stages-dependency.png differ