Release docker img

.github/workflows/release.yaml

@@ -50,3 +50,67 @@ jobs:
         user: __token__
         password: ${{ secrets.TEST_PYPI_API_TOKEN }}
         repository_url: https://test.pypi.org/legacy/
+
+  build-docker:
+    name: Build llm-foundry Release Docker Image
+    needs:
+    - code-quality
+    runs-on: mosaic-8wide
+    if: github.repository_owner == 'mosaicml'
+    steps:
+    - name: Checkout source
+      uses: actions/checkout@v3
+
+    - name: Setup Docker Buildx
+      uses: docker/setup-buildx-action@v2
+
+    - name: Login to DockerHub
+      uses: docker/login-action@v2
+      with:
+        username: ${{ secrets.DOCKER_HUB_USERNAME }}
+        password: ${{ secrets.DOCKER_HUB_PASSWORD }}
+
+    - name: Define Docker tags
+      id: define-tags
+      run: |
+        BRANCH_NAME="${{ github.ref_name }}"
+        TAG_NAME=$(echo "${BRANCH_NAME}" | sed 's/\//_/g')
+        echo "BRANCH_NAME=${BRANCH_NAME}" >> $GITHUB_ENV
+
+        echo "DOCKER_TAG=mosaicml/llm-foundry:release_${TAG_NAME}" >> $GITHUB_ENV
+        echo "AWS_DOCKER_TAG=mosaicml/llm-foundry:release_${TAG_NAME}_aws" >> $GITHUB_ENV
+        echo "LATEST_TAG=mosaicml/llm-foundry:release-latest" >> $GITHUB_ENV
+        echo "AWS_LATEST_TAG=mosaicml/llm-foundry:release_aws-latest" >> $GITHUB_ENV
+
+
+    - name: Build and push AWS Docker image
+      uses: docker/build-push-action@v3
+      with:
+        context: .
+        file: Dockerfile
+        push: true
+        tags: |
+          ${{ env.AWS_DOCKER_TAG }}
+          ${{ env.AWS_LATEST_TAG }}
+        build-args: |
+          BASE_IMAGE=mosaicml/pytorch:2.4.0_cu124-python3.11-ubuntu20.04-aws
+          BRANCH_NAME=${{ env.BRANCH_NAME }}
+          TE_COMMIT=901e5d2
+          DEP_GROUPS=[all]
+          KEEP_FOUNDRY=true
+
+    - name: Build and push Docker image
+      uses: docker/build-push-action@v3
+      with:
+        context: .
+        file: Dockerfile
+        push: true
+        tags: |
+          ${{ env.DOCKER_TAG }}
+          ${{ env.LATEST_TAG }}
+        build-args: |
+          BASE_IMAGE=mosaicml/pytorch:2.4.0_cu124-python3.11-ubuntu20.04
+          BRANCH_NAME=${{ env.BRANCH_NAME }}
+          TE_COMMIT=901e5d2
+          DEP_GROUPS=[all]
+          KEEP_FOUNDRY=true

Dockerfile

@@ -7,6 +7,7 @@ FROM $BASE_IMAGE
 ARG BRANCH_NAME
 ARG DEP_GROUPS
 ARG TE_COMMIT
+ARG KEEP_FOUNDRY=false
 
 ENV TORCH_CUDA_ARCH_LIST="8.0 8.6 8.7 8.9 9.0"
 
@@ -21,5 +22,9 @@ RUN NVTE_FRAMEWORK=pytorch CMAKE_BUILD_PARALLEL_LEVEL=4 MAX_JOBS=4 pip install g
 # Install and uninstall foundry to cache foundry requirements
 RUN git clone -b $BRANCH_NAME https://github.com/mosaicml/llm-foundry.git
 RUN pip install --no-cache-dir "./llm-foundry${DEP_GROUPS}"
-RUN pip uninstall -y llm-foundry
-RUN rm -rf llm-foundry
+
+# Conditionally uninstall llm-foundry and remove its directory
+RUN if [ "$KEEP_FOUNDRY" != "true" ]; then \
+      pip uninstall -y llm-foundry && \
+      rm -rf /llm-foundry; \
+    fi