.github/workflows/dataset_update.yml

name: WhisperKit Evals Dataset Update Workflow

on:
  schedule:
    - cron: "0 * * * *"
  workflow_dispatch:

jobs:
  update-datasets:
    runs-on: ubuntu-latest
    permissions:
      contents: write
    steps:
      - uses: actions/checkout@v4

      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      - name: Install dependencies
        run: |
          python -m pip install --upgrade pip
          pip install huggingface_hub requests

      - name: Upload relevant dashboard data for report generation
        uses: actions/upload-artifact@v4
        with:
          name: report_data
          path: |
            dashboard_data/performance_data.json
            dashboard_data/support_data.csv
            dashboard_data/version.json

      - name: Check dataset updates
        id: check_updates
        env:
          HF_TOKEN: ${{ secrets.HF_TOKEN }}
        run: python .github/scripts/check_dataset_update.py

      - name: Save workflow data
        run: |
          mkdir -p ./workflow_data
          echo "${{ steps.check_updates.outputs.has_updates }}" > ./workflow_data/has_updates.txt

      - name: Upload workflow data
        uses: actions/upload-artifact@v4
        with:
          name: workflow_data
          path: workflow_data/
          overwrite: true

      - name: Install full requirements
        if: steps.check_updates.outputs.has_updates == 'true'
        run: |
          pip install -r requirements.txt

      - name: Run generate script
        if: steps.check_updates.outputs.has_updates == 'true'
        run: |
          make update-performance-data

      - name: Commit and push if changed
        if: steps.check_updates.outputs.has_updates == 'true'
        env:
          GH_TOKEN: ${{ secrets.BOT_TOKEN }}
          GH_USERNAME: ${{ secrets.BOT_USERNAME }}
          HF_TOKEN: ${{ secrets.HF_TOKEN }}
        run: |
          git config --global user.email "[email protected]"
          git config --global user.name $GH_USERNAME

          git remote set-url origin https://x-access-token:${GH_TOKEN}@github.com/${GITHUB_REPOSITORY}.git

          git add .
          git commit -m "update dataset files" || echo "No changes to commit"
          git push -f
          git push -f https://HF_USERNAME:[email protected]/spaces/argmaxinc/whisperkit-benchmarks

  generate-report:
    needs: update-datasets
    runs-on: ubuntu-latest
    steps:
      - uses: actions/checkout@v4
        with:
          ref: main
          fetch-depth: 0

      - name: Set up Python
        uses: actions/setup-python@v5
        with:
          python-version: "3.11"

      - name: Download workflow data
        uses: actions/download-artifact@v4
        with:
          name: workflow_data
          path: workflow_data

      - name: Check updates status
        id: check
        run: |
          HAS_UPDATES=$(cat workflow_data/has_updates.txt)
          echo "has_updates=$HAS_UPDATES" >> $GITHUB_OUTPUT

      - name: Download report data
        if: steps.check.outputs.has_updates == 'true'
        uses: actions/download-artifact@v4
        with:
          name: report_data
          path: report_data

      - name: Install dependencies
        if: steps.check.outputs.has_updates == 'true'
        run: |
          python -m pip install --upgrade pip
          pip install pandas beautifulsoup4

      - name: Process report
        if: steps.check.outputs.has_updates == 'true'
        id: report
        run: python .github/scripts/process_report.py

      - name: Post to a Slack Channel
        if: steps.check.outputs.has_updates == 'true'
        id: slack_message
        uses: slackapi/[email protected]
        with:
          channel-id: ${{ secrets.SLACK_CHANNEL_ID }}
          payload: |
            ${{ steps.report.outputs.slack_message_payload }}
        env:
          SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}

      - name: Send Thread Message
        if: steps.check.outputs.has_updates == 'true'
        uses: slackapi/[email protected]
        with:
          channel-id: ${{ secrets.SLACK_CHANNEL_ID }}
          payload: |
            {
              "thread_ts": "${{ steps.slack_message.outputs.ts }}",
              "text": "${{ steps.report.outputs.performance_message }}"
            }
        env:
          SLACK_BOT_TOKEN: ${{ secrets.SLACK_BOT_TOKEN }}