CARRARO, FrancescoFrancescoCARRAROFONTE, SERGIOSERGIOFONTE2025-04-072025-04-072025http://hdl.handle.net/20.500.12386/37036https://doi.org/10.20371/INAF/TechRep/339The HyperLab project, funded in 2023, initially aimed at developing a mobile/web application for data analysis related to the SLab laboratory data at INAF-IAPS in Rome. The initial phase of the project focused on creating a user-friendly interface for analyzing and visualizing data from the SLab laboratory, significantly improving the efficiency of data analysis tasks. The project aims to fully leverage the powerful AI models developed by leading technology companies to significantly accelerate the creation of a functional product from scratch. The team's approach focuses on integrating pre-built models, adapting and customizing them to meet specific project requirements. This method enables small teams to effectively compete in the AI space, tapping into the collective expertise of advanced tech organizations. The model training process involved several key phases, including generating training data, creating tokens, and defining metrics to evaluate the quality of the training. The selected model, Google's T5-small, was chosen for its low resource requirements, making it ideal for an initial study phase. The project utilized containerization techniques to ensure portability, isolation, and efficiency. In summary, the HyperLab project represents an innovative and pragmatic approach to AI model training, leveraging advanced technologies to accelerate development and improve the quality of results.ELETTRONICOenTechnical reportINF/01 - INFORMATICA