Automation glossary

Fine-Tuning

Fine-tuning is further training a pre-trained AI model on your own examples so it performs a specific task or follows a specific style more consistently.

In plain language

It teaches an existing model a narrow skill by showing it many examples of the right output.

Example

A model is fine-tuned on thousands of past ticket categorizations so it labels new tickets the way the support team does.

Large Language Model (LLM) · Retrieval-Augmented Generation (RAG) · Classification · Evals

Common misconception

Fine-tuning is not the best way to add facts or keep knowledge current; retrieval is usually better for changing information.

Implementation guidance

Try prompting and retrieval first, use clean and representative examples, hold out a test set, and compare results with evals before switching.

Primary references

← Browse every glossary term