* 🔧 Update MkDocs for other languages. PR [#9734](https://github.com/tiangolo/fastapi/pull/9734) by [@tiangolo](https://github.com/tiangolo).
* 👷 Refactor Docs CI, run in multiple workers with a dynamic matrix to optimize speed. PR [#9732](https://github.com/tiangolo/fastapi/pull/9732) by [@tiangolo](https://github.com/tiangolo).

        with:
          python-version: ${{ matrix.python-version }}
          # Issue ref: https://github.com/actions/setup-python/issues/436
          # cache: "pip"
          # cache-dependency-path: pyproject.toml
      - uses: actions/cache@v4
        id: cache
        with:
          path: ${{ env.pythonLocation }}

* **Machine Learning**: it normally requires lots of "matrix" and "vector" multiplications. Think of a huge spreadsheet with numbers and multiplying all of them together at the same time.

        run: python ./scripts/docs.py update-languages
      - uses: actions/cache@v4
        with:
          key: mkdocs-cards-${{ matrix.lang }}-${{ github.ref }}
          path: docs/${{ matrix.lang }}/.cache
      - name: Build Docs
        run: python ./scripts/docs.py build-lang ${{ matrix.lang }}
      - uses: actions/upload-artifact@v3
        with:
          name: docs-site
          path: ./site/**

* **Maschinelles Lernen**: Normalerweise sind viele „Matrix“- und „Vektor“-Multiplikationen erforderlich. Stellen Sie sich eine riesige Tabelle mit Zahlen vor, in der Sie alle Zahlen gleichzeitig multiplizieren.