Merge branch 'vllm-project:main' into vllm-model

Author: SilverSoldier

.github/CODEOWNERS

@@ -0,0 +1,28 @@
+# See https://help.github.com/articles/about-codeowners/
+# for more info about CODEOWNERS file
+
+# These owners will be the default owners for everything in the repo
+# and will be requested for review when someone opens a pull request
+# unless a later match takes precedence
+* @joerunde @prashantgupta24
+
+# This lists cover the "core" components of vLLM-Spyre that require careful review
+/vllm_spyre @yannicks1 @tdoublep @nikolaospapandreou @sducouedic
+
+# TODO: separate code into different packages:
+#   /vllm_spyre/v1/worker/continuous_batching @nikolaospapandreou
+#   /vllm_spyre/v1/worker/static_batching ...
+
+# Tests
+/tests @rafvasq @prashantgupta24 @sducouedic
+
+# Examples
+/examples @yannicks1 @prashantgupta24 @sducouedic
+
+# Docs
+/docs @rafvasq
+.readthedocs.yaml @rafvasq
+mkdocs.yaml @rafvasq
+
+# CI
+/.github @joerunde @ckadner

.github/workflows/add_label_automerge.yml

@@ -1,21 +1,23 @@
 name: Add label on auto-merge enabled
+
 on:
-    pull_request_target:
-        types:
-            - auto_merge_enabled
+  pull_request_target:
+    types:
+      - auto_merge_enabled
+
 jobs:
-    add-label-on-auto-merge:
-        runs-on: ubuntu-latest
-        steps:
-            -   name: Add label
-                uses: actions/github-script@60a0d83039c74a4aee543508d2ffcb1c3799cdea # v7.0.1
-                with:
-                    script: |
-                        github.rest.issues.addLabels({
-                            owner: context.repo.owner,
-                            repo: context.repo.repo,
-                            issue_number: context.issue.number,
-                            labels: ['ready']
-                        })
-                env:
-                    GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
+  add-label-on-auto-merge:
+    runs-on: ubuntu-latest
+    steps:
+      - name: Add label
+        uses: actions/github-script@60a0d83039c74a4aee543508d2ffcb1c3799cdea # v7.0.1
+        with:
+          script: |
+            github.rest.issues.addLabels({
+              owner: context.repo.owner,
+              repo: context.repo.repo,
+              issue_number: context.issue.number,
+              labels: ['ready']
+            })
+        env:
+          GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}

.github/workflows/build-and-publish.yaml renamed to .github/workflows/build_and_publish.yaml

@@ -0,0 +1,106 @@
+name: "Build Docker"
+
+on:
+  workflow_dispatch:
+  push:
+    branches:
+      - "main"
+    paths-ignore:
+      - "**.md"
+  pull_request:
+    branches:
+      - "main"
+    paths-ignore:
+      - "**.md"
+  release:
+    types: [published]
+
+defaults:
+  run:
+    shell: bash
+
+jobs:
+  build:
+    runs-on: ubuntu-latest
+    env:
+      CACHE_IMAGE: "ghcr.io/vllm-project/vllm-spyre:build-cache"
+      QUAY_REPOSITORY: "quay.io/ibm-aiu/vllm-spyre"
+
+    permissions:
+      packages: write
+      contents: read
+
+    steps:
+    - name: "Checkout"
+      uses: actions/checkout@v4
+      # We need to fetch history and tags so that we can correctly version the
+      # vllm_spyre package, since it uses setuptools_scm to version based on
+      # git tags.
+      with:
+        fetch-depth: 0
+
+    - name: "Set up QEMU"
+      uses: docker/setup-qemu-action@v3
+
+    - name: "Set up Docker Buildx"
+      uses: docker/setup-buildx-action@v3
+
+    - name: "Log in to Github Container registry"
+      uses: docker/login-action@v3
+      if: github.event_name != 'pull_request'
+      with:
+        registry: ghcr.io
+        username: ${{ github.actor }}
+        password: ${{ secrets.GITHUB_TOKEN }}
+
+    - name: "Log in to quay"
+      uses: docker/login-action@v3
+      if: github.event_name != 'pull_request'
+      with:
+        registry: quay.io
+        username: ibm-aiu+aiu_pusher_bot
+        password: ${{ secrets.IBM_AIU_QUAY_TOKEN }}
+
+    - name: "Set build cache target"
+      run: |
+        # For push to `main` (PR merged), push a new cache image with all layers (cache-mode=max).
+        # For PR builds, use GitHub action cache which isolates cached layers by PR/branch.
+        # to optimize builds for subsequent pushes to the same PR/branch.
+        # Do not set a cache-to image for PR builds to not overwrite the `main` cache image and
+        # to not ping-pong cache images for two or more different PRs.
+        # Do not push cache images for each PR or multiple branches to not exceed GitHub package
+        # usage and traffic limitations.
+        # UPDATE 2024/02/26: GHA cache appears to have issues, cannot use `cache-to: gha,mode=min`
+        # if `cache-from: reg...,mode=max` but `cache-to: gha,mode=max` takes longer than uncached
+        # build and exhausts GHA cache size limits, so use cache `type=inline` (no external cache).
+        if [ "${{ github.event_name }}" == "pull_request" ]
+        then
+            #CACHE_TO="type=gha,mode=min"
+            CACHE_TO="type=inline"
+        else
+            CACHE_TO="type=registry,ref=${{ env.CACHE_IMAGE }},mode=max"
+        fi
+        echo "CACHE_TO=$CACHE_TO" >> "$GITHUB_ENV"
+
+    - name: "Generate tags"
+      id: meta
+      uses: docker/metadata-action@v5
+      with:
+        images: |
+          ${{ env.QUAY_REPOSITORY }}
+        tags: |
+          type=raw,value=latest.amd64
+          type=semver,pattern={{version}}.amd64
+
+    - name: "Build and push"
+      uses: docker/build-push-action@v5
+      with:
+        context: .
+        file: docker/Dockerfile.amd64
+        target: release
+        provenance: false
+        tags: ${{ steps.meta.outputs.tags }}
+        cache-from: type=registry,ref=${{ env.CACHE_IMAGE }}
+        cache-to: ${{ env.CACHE_TO }}
+        push: ${{ github.event_name != 'pull_request' }}
+        platforms: linux/amd64

.github/workflows/build_docker.yml

@@ -1,19 +1,16 @@
-name: Lint GitHub Actions workflows
+name: Lint GitHub Actions
+
 on:
-  push:
-    branches:
-      - "main"
-    paths:
-      - '.github/workflows/*.ya?ml'
-      - '.github/workflows/actionlint.*'
-      - '.github/workflows/matchers/actionlint.json'
+  # Don't use pull_request.paths filter since this workflow is required for
+  # all pull requests on main irrespective of file type or location
   pull_request:
+    branches:
+      - main
+  push:
     branches:
       - "main"
     paths:
       - '.github/workflows/*.ya?ml'
-      - '.github/workflows/actionlint.*'
-      - '.github/workflows/matchers/actionlint.json'
 
 env:
   LC_ALL: en_US.UTF-8
@@ -26,7 +23,7 @@ permissions:
   contents: read
 
 jobs:
-  actionlint:
+  lint-actions:
     runs-on: ubuntu-latest
     steps:
       - name: "Checkout"
@@ -35,6 +32,8 @@ jobs:
           fetch-depth: 0
 
       - name: "Run actionlint"
-        run: |
-          echo "::add-matcher::.github/workflows/matchers/actionlint.json"
-          tools/actionlint.sh -color
+        uses: raven-actions/actionlint@v2
+        with:
+          matcher: true
+          cache: false
+          fail-on-error: true

.github/workflows/actionlint.yml renamed to .github/workflows/lint_actions.yml

@@ -1,34 +1,22 @@
-name: lint code
+name: Lint Code
 
 on:
-  # Trigger the workflow on push or pull request,
-  # but only for the main branch
+  # Don't use pull_request.paths filter since this workflow is required for
+  # all pull requests on main irrespective of file type or location
+  pull_request:
+    branches:
+      - main
   push:
     branches:
       - main
     paths:
       - "**/*.py"
       - pyproject.toml
-      - requirements-lint.txt
       - .github/workflows/matchers/ruff.json
-      - .github/workflows/ruff.yml
-  pull_request:
-    branches:
-      - main
-    # This workflow is only relevant when one of the following files changes.
-    # However, we have github configured to expect and require this workflow
-    # to run and pass before github with auto-merge a pull request. Until github
-    # allows more flexible auto-merge policy, we can just run this on every PR.
-    # It doesn't take that long to run, anyway.
-    #paths:
-    #  - "**/*.py"
-    #  - pyproject.toml
-    #  - requirements-lint.txt
-    #  - .github/workflows/matchers/ruff.json
-    #  - .github/workflows/ruff.yml
+      - .github/workflows/lint_code.yml
 
 jobs:
-  ruff:
+  lint-code:
     runs-on: ubuntu-latest
     strategy:
       matrix:

.github/workflows/lint_code.yml

@@ -1,19 +1,21 @@
-name: Lint documentation
+name: Lint Docs
 
 on:
-  push:
+  # Trigger the workflow on push or pull request, but only for the main branch.
+  # Don't use pull_request.paths filter since this workflow is required for
+  # all pull requests on main irrespective of file type or location.
+  pull_request:
     branches:
       - main
-    paths:
-      - "docs/**"
-  pull_request:
+  push:
     branches:
       - main
     paths:
+      - "**/*.md"
       - "docs/**"
 
 jobs:
-  doc-lint:
+  lint-docs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
@@ -27,4 +29,4 @@ jobs:
       - name: Install dependencies
         run: uv sync --frozen --only-group lint
       - name: Lint docs
-        run: tools/doc-lint.sh
+        run: tools/lint_docs.sh

.github/workflows/doc-lint.yml renamed to .github/workflows/lint_docs.yml

@@ -1,17 +1,17 @@
-name: Lint shell scripts
+name: Lint Scripts
+
 on:
-  push:
-    branches:
-      - "main"
-    paths:
-      - '**/*.sh'
-      - '.github/workflows/shellcheck.yml'
+  # Don't use pull_request.paths filter since this workflow is required for
+  # all pull requests on main irrespective of file type or location
   pull_request:
     branches:
-      - "main"
+      - main
+  push:
+    branches:
+      - main
     paths:
       - '**/*.sh'
-      - '.github/workflows/shellcheck.yml'
+      - '.github/workflows/lint_scripts.yml'
 
 env:
   LC_ALL: en_US.UTF-8
@@ -24,7 +24,7 @@ permissions:
   contents: read
 
 jobs:
-  shellcheck:
+  lint-scripts:
     runs-on: ubuntu-latest
     steps:
       - name: "Checkout"
@@ -34,4 +34,4 @@ jobs:
 
       - name: "Check shell scripts"
         run: |
-          tools/shellcheck.sh
+          tools/lint_scripts.sh