Forecasting solar energy is essential for efficient integration of photovoltaics (PV) into electricity grids and optimally managing renewable energy resources. As PV becomes an increasingly significant component of the energy landscape, accurate predictions of its availability and output are required for efficient integration into energy systems and ensuring a steady supply of electricity. Artificial intelligence (AI)-based PV forecasts are particularly valuable due to their ability to accurately predict PV generation by extracting complex relationships between different variables. However, their performance is heavily influenced by the selection of hyperparameters, such as learning rate, network architecture, and regularisation terms, which significantly affect the model's structure, learning efficiency, and forecasting accuracy. This chapter focuses on the importance of model tuning in AI-driven PV forecasting and explains how selecting the right hyperparameters can enhance forecast accuracy and support more reliable solar energy integration.