AI & LLMs

Fine-Tuning

Continuing to train a pretrained model on your data to specialize its behavior.

Fine-tuning adapts a base model to a task, domain, or style by training further on curated examples. Parameter-efficient methods like LoRA update only small adapter weights, making it cheap. Often compared against prompting and RAG as ways to change model behavior.