The aim of this milestone is to establish a robust foundation for upcoming experiments. The code should prioritize maintainability, scalability, and flexibility to accommodate the execution of experiments with minimal modifications. It will include a basic training pipeline equipped with validation and evaluation capabilities, catering to both toy datasets and the proposed dataset. Additionally, fundamental metrics for assessment will be integrated.