chore: build manylinux2_28 builder image (#435)

Author: guocuimi

.github/workflows/publish_manylinux_2_28_image.yml

@@ -0,0 +1,60 @@
+name: Publish manylinux2_28 image
+on:
+  workflow_dispatch:
+    inputs:
+      push:
+        description: 'Push to Docker Hub'
+        required: false
+        default: 'true'
+
+  # Schedule the workflow to run at 9:00 (UTC) every month.
+  schedule:
+    # Minute[0,59] Hour[0,23] Day of month[1,31] Month[1,12] Day of week[0,6] (Sunday=0)
+    - cron: '0 9 1 * *'
+
+env:
+  # Tells where to store caches.
+  CI_CACHE_DIR: ${{ github.workspace }}/../../ci_cache
+
+jobs:
+  publish_manylinux_2_28_image:
+    strategy:
+      fail-fast: false
+      matrix:
+        cuda: ["12.6", "12.8"]
+    runs-on: [self-hosted, linux, build]
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+        with:
+          cache-image: false
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+        with:
+          cache-binary: false
+
+      - name: Login to Docker Hub
+        uses: docker/login-action@v3
+        with:
+          username: ${{ secrets.DOCKER_HUB_USER }}
+          password: ${{ secrets.DOCKER_HUB_TOKEN }}
+
+      - name: Create cache directory
+        run: mkdir -p $CI_CACHE_DIR/.buildx-cache
+
+      - name: Build base for cuda ${{ matrix.cuda }}
+        uses: docker/build-push-action@v5
+        with:
+          context: ./docker
+          file: ./docker/Dockerfile.manylinux2_28
+          push: ${{ inputs.push || 'true' }}
+          cache-from: type=local,src=$CI_CACHE_DIR/.buildx-cache
+          cache-to: type=local,dest=$CI_CACHE_DIR/.buildx-cache
+          build-args: |
+            CUDA_VERSION=${{ matrix.cuda }}
+          tags: |
+            vectorchai/scalellm_manylinux2_28:cuda${{ matrix.cuda }}

.github/workflows/publish_manylinux_image.yml

@@ -7,11 +7,6 @@ on:
         required: false
         default: 'true'
 
-  # Schedule the workflow to run at 9:00 (UTC) every month.
-  schedule:
-    # Minute[0,59] Hour[0,23] Day of month[1,31] Month[1,12] Day of week[0,6] (Sunday=0)
-    - cron: '0 9 1 * *'
-
 env:
   # Tells where to store caches.
   CI_CACHE_DIR: ${{ github.workspace }}/../../ci_cache
@@ -26,7 +21,7 @@ jobs:
     steps:
       - name: Checkout repository
         uses: actions/checkout@v4
-      
+
       - name: Set up QEMU
         uses: docker/setup-qemu-action@v3
         with:
@@ -58,4 +53,3 @@ jobs:
             CUDA_VERSION=${{ matrix.cuda }}
           tags: |
             vectorchai/scalellm_manylinux:cuda${{ matrix.cuda }}
-

docker/Dockerfile.manylinux

@@ -1,5 +1,5 @@
 ARG CUDA_VERSION=12.1
-FROM pytorch/manylinux-builder:cuda${CUDA_VERSION} as base
+FROM pytorch/manylinux-builder:cuda${CUDA_VERSION} AS base
 
 LABEL maintainer="[email protected]"
 ENV DEBIAN_FRONTEND=noninteractive
@@ -47,4 +47,4 @@ RUN curl https://sh.rustup.rs -sSf | sh -s -- -y
 RUN chmod -R a+w ${RUSTUP_HOME} ${CARGO_HOME}
 RUN rustup --version; cargo --version; rustc --version
 
-CMD ["bash"]
+CMD ["bash"]

docker/Dockerfile.manylinux2_28

@@ -0,0 +1,37 @@
+ARG CUDA_VERSION=12.6
+FROM pytorch/manylinux2_28-builder:cuda${CUDA_VERSION} AS base
+
+LABEL maintainer="[email protected]"
+ENV DEBIAN_FRONTEND=noninteractive
+
+ENV LC_ALL=en_US.UTF-8
+ENV LANG=en_US.UTF-8
+ENV LANGUAGE=en_US.UTF-8
+
+# Show versions of gcc, g++, nvcc
+RUN gcc --version; g++ --version; nvcc --version
+
+# Install cmake
+ARG CMAKE_VERSION=3.29.3
+COPY ./common/install_cmake.sh install_cmake.sh
+RUN if [ -n "${CMAKE_VERSION}" ]; then bash ./install_cmake.sh; fi
+RUN rm install_cmake.sh
+RUN cmake --version
+
+# Install ccache
+ARG CCACHE_VERSION=4.8.3
+COPY ./common/install_ccache.sh install_ccache.sh
+RUN if [ -n "${CCACHE_VERSION}" ]; then bash ./install_ccache.sh; fi
+RUN rm install_ccache.sh
+RUN ccache --version
+
+# Install rust
+ENV RUSTUP_HOME=/usr/local/rustup
+ENV CARGO_HOME=/usr/local/cargo
+ENV PATH=/usr/local/cargo/bin:$PATH
+RUN curl https://sh.rustup.rs -sSf | sh -s -- -y
+# give everyone permission to use rust
+RUN chmod -R a+w ${RUSTUP_HOME} ${CARGO_HOME}
+RUN rustup --version; cargo --version; rustc --version
+
+CMD ["bash"]