Summary
In this chapter, we explored two model customization techniques, fine-tuning and continued pre-training, the need to customize a model, and understood the concepts behind fine-tuning and continued pre-training. Further, we prepared our dataset, created a custom model, evaluated the model, and performed inference.
Lastly, we discussed some of the guidelines and best practices you need to consider when customizing your FM.
In the next chapter, we’re going to uncover the power of RAG in solving real-world business problems by using an external data source. We will delve into the various use cases and sample architectures and implement RAG with Amazon Bedrock.
