Add workflow to run SDGym multi-table benchmark monthly and publish results

.github/workflows/run_benchmark.yml

@@ -1,31 +1,77 @@
 name: Run SDGym Benchmark
 
 on:
-  workflow_dispatch:
-  schedule:
-    - cron: '0 5 1 * *'
+  workflow_call:
+    inputs:
+      modality:
+        required: true
+        type: string
+      sdgym_ref:
+        required: false
+        type: string
+        default: issue-516-add-workflows
+    secrets:
+      SDV_ENTERPRISE_USERNAME:
+        required: true
+      SDV_ENTERPRISE_LICENSE_KEY:
+        required: true
+      GCP_SERVICE_ACCOUNT_JSON:
+        required: true
+      AWS_ACCESS_KEY_ID:
+        required: true
+      AWS_SECRET_ACCESS_KEY:
+        required: true
+      SLACK_TOKEN:
+        required: true
 
 jobs:
   run-sdgym-benchmark:
     runs-on: ubuntu-latest
+
     steps:
-    - uses: actions/checkout@v4
-      with:
-        fetch-depth: 0
-    - name: Set up latest Python
-      uses: actions/setup-python@v5
-      with:
-        python-version-file: 'pyproject.toml'
-    - name: Install dependencies
-      run: |
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v5
+        with:
+          python-version-file: "pyproject.toml"
+
+      - name: Install dependencies
+        env:
+          USERNAME: ${{ secrets.SDV_ENTERPRISE_USERNAME }}
+          LICENSE_KEY: ${{ secrets.SDV_ENTERPRISE_LICENSE_KEY }}
+        run: |
+          python -m venv venv
+          source venv/bin/activate
+
           python -m pip install --upgrade pip
-          python -m pip install --no-cache-dir -e .[dev]
-
-    - name: Run SDGym Benchmark
-      env:
-        SLACK_TOKEN: ${{ secrets.SLACK_TOKEN }}
-        AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
-        AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
-        AWS_DEFAULT_REGION: ${{ secrets.AWS_REGION }}
-
-      run: invoke run-sdgym-benchmark
+          python -m pip install sdv-installer
+          python -c "
+          from sdv_installer.installation.installer import install_packages
+          install_packages(
+              username='${USERNAME}',
+              license_key='${LICENSE_KEY}',
+              package='sdv-enterprise',
+          )
+          "
+          python -m pip install "sdgym[all] @ git+https://github.com/sdv-dev/SDGym.git@${{ inputs.sdgym_ref }}"
+
+          echo "VIRTUAL_ENV=$(pwd)/venv" >> $GITHUB_ENV
+          echo "$(pwd)/venv/bin" >> $GITHUB_PATH
+
+      - name: Run SDGym Benchmark
+        env:
+          GCP_SERVICE_ACCOUNT_JSON: ${{ secrets.GCP_SERVICE_ACCOUNT_JSON }}
+          GCP_PROJECT_ID: ${{ secrets.GCP_PROJECT_ID }}
+          GCP_ZONE: ${{ secrets.GCP_ZONE }}
+          SDV_ENTERPRISE_USERNAME: ${{ secrets.SDV_ENTERPRISE_USERNAME }}
+          SDV_ENTERPRISE_LICENSE_KEY: ${{ secrets.SDV_ENTERPRISE_LICENSE_KEY }}
+          AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
+          SLACK_TOKEN: ${{ secrets.SLACK_TOKEN }}
+        run: |
+          export CREDENTIALS_FILEPATH=$(python -c "from sdgym._benchmark.credentials_utils import create_credentials_file; print(create_credentials_file())")
+          invoke run-sdgym-benchmark --modality "${{ inputs.modality }}"
+          rm -f "$CREDENTIALS_FILEPATH"

.github/workflows/run_benchmark_multi_table.yaml

@@ -0,0 +1,16 @@
+name: Run SDGym Benchmark Multi-Table
+
+on:
+  workflow_dispatch:
+  schedule:
+    - cron: "0 5 1 * *"
+  push:
+    branches:
+      - issue-516-add-workflows
+
+jobs:
+  call-run-sdgym-benchmark:
+    uses: ./.github/workflows/run_benchmark.yml
+    with:
+      modality: multi_table
+    secrets: inherit

.github/workflows/run_benchmark_single_table.yml

@@ -0,0 +1,16 @@
+name: Run SDGym Benchmark Single-Table
+
+on:
+  workflow_dispatch:
+  schedule:
+    - cron: "0 5 1 * *"
+  push:
+    branches:
+      - issue-516-add-workflows
+
+jobs:
+  call-run-sdgym-benchmark:
+    uses: ./.github/workflows/run_benchmark.yml
+    with:
+      modality: single_table
+    secrets: inherit

.github/workflows/upload_benchmark_results.yml

@@ -1,91 +1,107 @@
-name: Upload SDGym Benchmark results
+name: Upload SDGym Benchmark Results
 
 on:
-  workflow_run:
-    workflows: ["Run SDGym Benchmark"]
-    types:
-      - completed
-  workflow_dispatch:
-  schedule:
-    - cron: '0 6 * * *'
+  workflow_call:
+    inputs:
+      modality:
+        description: "Benchmark modality to upload"
+        required: true
+        type: string
+    secrets:
+      PYDRIVE_TOKEN:
+        required: true
+      AWS_ACCESS_KEY_ID:
+        required: true
+      AWS_SECRET_ACCESS_KEY:
+        required: true
+      GH_TOKEN:
+        required: true
+      SLACK_TOKEN:
+        required: true
 
 jobs:
   upload-sdgym-benchmark:
     runs-on: ubuntu-latest
 
     steps:
-    - uses: actions/checkout@v4
-      with:
-        fetch-depth: 0
-
-    - name: Set up latest Python
-      uses: actions/setup-python@v5
-      with:
-        python-version-file: 'pyproject.toml'
-
-    - name: Install dependencies
-      run: |
-        python -m pip install --upgrade pip
-        python -m pip install --no-cache-dir -e .[dev]
-
-    - name: Upload SDGym Benchmark
-      env:
-        PYDRIVE_TOKEN: ${{ secrets.PYDRIVE_TOKEN }}
-        AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
-        AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
-        GITHUB_LOCAL_RESULTS_DIR: ${{ runner.temp }}/sdgym-leaderboard-files
-      run: |
-        invoke upload-benchmark-results
-        echo "GITHUB_LOCAL_RESULTS_DIR=$GITHUB_LOCAL_RESULTS_DIR" >> $GITHUB_ENV
-
-    - name: Prepare files for commit
-      if: env.SKIP_UPLOAD != 'true'
-      run: |
-        mkdir pr-staging
-        echo "Looking for files in: $GITHUB_LOCAL_RESULTS_DIR"
-        ls -l "$GITHUB_LOCAL_RESULTS_DIR"
-        for f in "$GITHUB_LOCAL_RESULTS_DIR"/*; do
-          if [ -f "$f" ]; then
-            base=$(basename "$f")
-            cp "$f" "pr-staging/${base}"
-          fi
-        done
-
-        echo "Files staged for PR:"
-        ls -l pr-staging
-
-    - name: Checkout target repo (sdv-dev.github.io)
-      if: env.SKIP_UPLOAD != 'true'
-      run: |
-        git clone https://github.com/sdv-dev/sdv-dev.github.io.git target-repo
-        cd target-repo
-        git checkout gatsby-home
-
-    - name: Copy results and commit
-      if: env.SKIP_UPLOAD != 'true'
-      env:
-        GH_TOKEN: ${{ secrets.GH_TOKEN }}
-        FOLDER_NAME: ${{ env.FOLDER_NAME }}
-      run: |
-        cp pr-staging/* target-repo/assets/sdgym-leaderboard-files/
-        cd target-repo
-        git checkout gatsby-home
-        git config --local user.name "github-actions[bot]"
-        git config --local user.email "41898282+github-actions[bot]@users.noreply.github.com"
-        git add assets/
-        git commit -m "Upload SDGym Benchmark Results ($FOLDER_NAME)" || echo "No changes to commit"
-        git remote set-url origin https://x-access-token:${GH_TOKEN}@github.com/sdv-dev/sdv-dev.github.io.git
-        git push origin gatsby-home
-        COMMIT_HASH=$(git rev-parse HEAD)
-        COMMIT_URL="https://github.com/sdv-dev/sdv-dev.github.io/commit/${COMMIT_HASH}"
-        echo "Commit URL: $COMMIT_URL"
-        echo "COMMIT_URL=$COMMIT_URL" >> $GITHUB_ENV
-
-    - name: Send Slack notification
-      if: env.SKIP_UPLOAD != 'true'
-      env:
-        SLACK_TOKEN: ${{ secrets.SLACK_TOKEN }}
-      run: |
-        invoke notify-sdgym-benchmark-uploaded \
-          --folder-name "$FOLDER_NAME" \
-          --commit-url "$COMMIT_URL"
+      - uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Set up latest Python
+        uses: actions/setup-python@v5
+        with:
+          python-version-file: "pyproject.toml"
+
+      - name: Install dependencies
+        run: |
+          python -m pip install --upgrade pip
+          python -m pip install --no-cache-dir -e .[dev]
+
+      - name: Upload SDGym Benchmark
+        env:
+          PYDRIVE_TOKEN: ${{ secrets.PYDRIVE_TOKEN }}
+          AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
+          GITHUB_LOCAL_RESULTS_DIR: ${{ runner.temp }}/sdgym-leaderboard-files
+        run: |
+          invoke upload-benchmark-results --modality "${{ inputs.modality }}"
+          echo "GITHUB_LOCAL_RESULTS_DIR=$GITHUB_LOCAL_RESULTS_DIR" >> $GITHUB_ENV
+
+      - name: Prepare files for commit
+        if: env.SKIP_UPLOAD != 'true'
+        run: |
+          set -euo pipefail
+          mkdir -p pr-staging
+
+          echo "Looking for files in: $GITHUB_LOCAL_RESULTS_DIR"
+          ls -l "$GITHUB_LOCAL_RESULTS_DIR" || true
+
+          shopt -s nullglob
+          for f in "$GITHUB_LOCAL_RESULTS_DIR"/*; do
+            [ -f "$f" ] && cp "$f" "pr-staging/$(basename "$f")"
+          done
+
+          echo "Files staged for PR:"
+          ls -l pr-staging || true
+
+      - name: Checkout target repo (sdv-dev.github.io)
+        if: env.SKIP_UPLOAD != 'true'
+        run: |
+          git clone https://github.com/sdv-dev/sdv-dev.github.io.git target-repo
+          cd target-repo
+          git checkout gatsby-home
+
+      - name: Copy results and commit
+        if: env.SKIP_UPLOAD != 'true'
+        env:
+          GH_TOKEN: ${{ secrets.GH_TOKEN }}
+          FOLDER_NAME: ${{ env.FOLDER_NAME }}
+        run: |
+          set -euo pipefail
+
+          cp -f pr-staging/* target-repo/assets/sdgym-leaderboard-files/ || true
+          cd target-repo
+
+          git config --local user.name "github-actions[bot]"
+          git config --local user.email "41898282+github-actions[bot]@users.noreply.github.com"
+
+          git add assets/
+          git commit -m "Upload SDGym Benchmark Results ($FOLDER_NAME) - Modality: ${{ inputs.modality }}" || echo "No changes to commit"
+
+          git remote set-url origin "https://x-access-token:${GH_TOKEN}@github.com/sdv-dev/sdv-dev.github.io.git"
+          git push origin gatsby-home
+
+          COMMIT_HASH=$(git rev-parse HEAD)
+          COMMIT_URL="https://github.com/sdv-dev/sdv-dev.github.io/commit/${COMMIT_HASH}"
+          echo "COMMIT_URL=$COMMIT_URL" >> $GITHUB_ENV
+
+      - name: Send Slack notification
+        if: env.SKIP_UPLOAD != 'true'
+        env:
+          SLACK_TOKEN: ${{ secrets.SLACK_TOKEN }}
+        run: |
+          invoke notify-sdgym-benchmark-uploaded \
+            --folder-name "$FOLDER_NAME" \
+            --commit-url "$COMMIT_URL" \
+            --modality "${{ inputs.modality }}"

.github/workflows/upload_benchmark_results_multi_table.yml

@@ -0,0 +1,19 @@
+name: Upload SDGym Multi-Table Benchmark results
+
+on:
+  workflow_run:
+    workflows: ["Run SDGym Benchmark Multi-Table"]
+    types: [completed]
+  workflow_dispatch:
+  schedule:
+    - cron: "0 6 * * *"
+  push:
+    branches:
+      - issue-516-add-workflows
+
+jobs:
+  call-upload-benchmark-results:
+    uses: ./.github/workflows/upload_benchmark_results.yml
+    with:
+      modality: multi_table
+    secrets: inherit

.github/workflows/upload_benchmark_results_single_table.yml

@@ -0,0 +1,16 @@
+name: Upload SDGym Single-Table Benchmark results
+
+on:
+  workflow_run:
+    workflows: ["Run SDGym Benchmark Single-Table"]
+    types: [completed]
+  workflow_dispatch:
+  schedule:
+    - cron: "0 6 * * *"
+
+jobs:
+  call-upload-benchmark-results:
+    uses: ./.github/workflows/upload_benchmark_results.yml
+    with:
+      modality: single_table
+    secrets: inherit

sdgym/_benchmark/benchmark.py

@@ -204,7 +204,7 @@ def _get_user_data_script(
         log "======== Install Dependencies =========="
         pip install --upgrade pip
         {sdv_install}
-        pip install "sdgym[all] @ git+https://github.com/sdv-dev/SDGym.git@issue-515-_benchmark_multi_table_compute_gcp"
+        pip install "sdgym[all] @ git+https://github.com/sdv-dev/SDGym.git@issue-516-add-workflows"
 
         {gpu_block}
 
@@ -428,7 +428,7 @@ def _benchmark_single_table_compute_gcp(
     limit_dataset_size=False,
     compute_quality_score=True,
     compute_diagnostic_score=True,
-    compute_privacy_score=True,
+    compute_privacy_score=False,
     sdmetrics=None,
     timeout=None,
 ):

sdgym/_benchmark/config_utils.py

@@ -118,6 +118,6 @@ def validate_compute_config(config):
 
 
 def _make_instance_name(prefix):
-    day = datetime.now(timezone.utc).strftime('%Y_%m_%d_%H:%M')
+    day = datetime.now(timezone.utc).strftime('%Y%m%d-%H%M')
     suffix = uuid.uuid4().hex[:6]
     return f'{prefix}-{day}-{suffix}'