fine-tuning

Process of adapting a pretrained machine-learning model to a specific task using additional, task-relevant data.