This folder contains information to get you started with prompt engineering on Amazon Bedrock.
Prompt engineering is the practice of optimizing the quality and performance of your foundation model's response to your request. Prompt engineering may involve:
- Word choice
- Phrasing
- Providing examples (few-shot learning)
- Use of line breaks and content separators
- Following established formats that align with how the model was trained
- Use of stop sequences to help the model know when it should stop generating text
While general best practices for prompt engineering are still emerging, it is best to understand how each model works. Some prompts may be portable between providers, but many won't. You will likely need to alter and optimize your prompts if you want to try different providers.
You can learn more about prompt engineering in the prompt engineering basics lab.
Below are prompt engineering guides provided by some of Amazon Bedrock's model providers:
We welcome community contributions! Please ensure your sample aligns with AWS best practices, and please update the Contents section of this README file with a link to your sample, along with a description.