Fine-Tuning
Working knowledgeTechniques and Architectures
Also called: Fine Tuning, Finetuning
Fine-tuning is the process of taking a pre-trained foundation model and updating its parameters by training it on a smaller, domain-specific dataset. It adapts a general-purpose model to a specialised task, medical report writing or legal contract analysis, teaching it the terminology and tone of that domain at a fraction of the cost of training a model from scratch.
In practice
Fine-tuning is how engineering teams turn a general model into something that fits a proprietary workflow, and it is the common alternative to paying a vendor to build a bespoke one. The diligence question is who owns the resulting weights and the dataset behind them, the company or the supplier who did the work.