Fine-tuning

Process of specializing an AI model on domain-specific data.

Fine-tuning is the practice of taking a pre-trained model (e.g., GPT, LLaMA) and re-training it on a specific dataset to improve performance in a particular domain. It is more expensive than prompt engineering but more powerful for repetitive and specialized use cases.