Worked on the DS4SD/docling repository to enhance documentation for Mixtral text generation examples by implementing a task parameter within the HuggingFaceEndpoint. This addition allows users to specify the model task directly, improving clarity and usability for those integrating machine learning workflows. The work focused on Python and leveraged data science and machine learning expertise to ensure the documentation accurately reflected endpoint usage. All changes were committed with DCO-compliant sign-off to maintain license and author integrity. The update aimed to streamline onboarding, reduce ambiguity, and support adoption by providing clearer guidance for users working with text generation models.