While RAG is the most common way to focus a foundation model on material outside its training set, Bharani Subramaniam and I now finish our current batch of patterns by examining our experience with Fine Tuning where we learned to prioritize curating high-quality data.