[vllm test] add vllm.yml and additional package (#160698)

Pull Request resolved: https://github.com/pytorch/pytorch/pull/160698
Approved by: https://github.com/huydhn
ghstack dependencies: #160116

.github/pytorch-probot.yml

@@ -27,6 +27,7 @@ ciflow_push_tags:
 - ciflow/trunk
 - ciflow/unstable
 - ciflow/xpu
+- ciflow/vllm
 - ciflow/torchbench
 - ciflow/op-benchmark
 - ciflow/pull

.github/workflows/_linux-build.yml

@@ -96,6 +96,13 @@ on:
         required: false
         type: string
         default: ""
+      build-external-packages:
+        description: |
+          If set, the build external packages and saves their wheels as artifacts
+          use command separated list of packages to build ex: 'vllm,transformers'.
+        required: false
+        type: string
+        default: ""
 
     secrets:
       HUGGING_FACE_HUB_TOKEN:
@@ -356,6 +363,26 @@ jobs:
           END_TIME=$(date +%s)
           echo "build_time=$((END_TIME - START_TIME))" >> "$GITHUB_OUTPUT"
 
+      - name: Build external packages
+        id: build-external-packages
+        if: inputs.build-external-packages != '' &&  steps.build.outcome != 'skipped'
+        uses: ./.github/actions/build-external-packages
+        with:
+          build-targets: ${{ inputs.build-external-packages }}
+          docker-image: ${{ steps.calculate-docker-image.outputs.docker-image }}
+          cuda-arch-list: ${{ inputs.cuda-arch-list }}
+          output-dir: external
+
+      - name: Move external packages to dist
+        if: steps.build-external-packages.outputs.output_dir != '' && steps.build-external-packages.outcome != 'skipped'
+        shell: bash
+        run: |
+          src="${{ steps.build-external-packages.outputs.output_dir }}"
+          if [ -d "$src" ]; then
+            mkdir -p "dist/$(dirname "$src")"
+            mv "$src" "dist/$(dirname "$src")/"
+          fi
+
       - name: Stop monitoring script
         if: ${{ always() && steps.monitor-script.outputs.monitor-script-pid }}
         shell: bash

.github/workflows/vllm.yml

@@ -0,0 +1,45 @@
+name: vllm-test
+
+on:
+  push:
+    tags:
+      - ciflow/vllm/*
+  workflow_dispatch:
+
+concurrency:
+  group: ${{ github.workflow }}-${{ github.event.pull_request.number || github.ref_name }}-${{ github.ref_type == 'branch' && github.sha }}-${{ github.event_name == 'workflow_dispatch' }}
+  cancel-in-progress: true
+
+permissions:
+  id-token: write
+  contents: read
+
+jobs:
+  get-label-type:
+    name: get-label-type
+    uses: pytorch/pytorch/.github/workflows/_runner-determinator.yml@main
+    if: ${{ (github.event_name != 'schedule' || github.repository == 'pytorch/pytorch') && github.repository_owner == 'pytorch' }}
+    with:
+      triggering_actor: ${{ github.triggering_actor }}
+      issue_owner: ${{ github.event.pull_request.user.login || github.event.issue.user.login }}
+      curr_branch: ${{ github.head_ref || github.ref_name }}
+      curr_ref_type: ${{ github.ref_type }}
+      opt_out_experiments: lf
+
+  torch-build-sm89:
+    name: sm89-vllm-test
+    uses: ./.github/workflows/_linux-build.yml
+    needs: get-label-type
+    with:
+      build-additional-packages: "vision audio torchao"
+      build-external-packages: "vllm"
+      build-environment: linux-jammy-cuda12.8-py3.12-gcc11-sm89
+      docker-image-name: ci-image:pytorch-linux-jammy-cuda12.8-cudnn9-py3.12-gcc11-vllm
+      cuda-arch-list: '8.9'
+      runner: linux.24xlarge.memory
+      test-matrix: |
+        { include: [
+          { config: "vllm_basic_correctness_test", shard: 1, num_shards: 1, runner: "${{ needs.get-label-type.outputs.label-type }}linux.g6.4xlarge.experimental.nvidia.gpu"  },
+          { config: "vllm_basic_models_test", shard: 1, num_shards: 1, runner: "${{ needs.get-label-type.outputs.label-type }}linux.g6.4xlarge.experimental.nvidia.gpu" },
+        ]}
+    secrets: inherit