Fine-Tuning

Fine-Tuning is the act of continuing to train a pre-trained model on a new dataset, for a different problem or to improve performance. It's a key part of the process of Transfer Learning.

In the LLM Training Pipeline, post-training techniques include Supervised Fine-Tuning and preference tuning (Raschka, 2026).

References

Sebastian Raschka. Build a Reasoning Model. Manning Publications, Erscheinungsort nicht ermittelbar, 2026. ISBN 978-1-63343-467-7. ↩