Effective LLM Prompting skills

Apply best practices for crafting effective prompts to LLM are key for effective prompting to LLM. Here are the best practices.

  • Role assignment: specifying a role for the LLM lead to a more helpful response. You can even apply multiple roles to LLM
  • Specificity and context: by providing context about your role and the situation and highlighting your specific concerns, you're guiding the LLM to provide specific and relevant information.
  • Feedback to LLM:  You can effectively analyze your prompt based on the LLM response, and identify weaknesses, and points out the targeted improvements for better prompt.
For example, you are a software developer and want LLM to review and debug the code errors.

Your prompt can be:

You are a senior Python developer and tester, I'm a junior Python developer working on a program to calculate daily interest rates. I'm encountering an error when running my code. Could you help me debug it? Here's a snippet of the code: "..."', and error throws at "....". The 'rate' variable is a number and the error message is "....". I'm particularly concerned about handling cases where the 'rate' variable might be invalid or missing. Could you also suggest ways to improve the code's clarity and documentation for future reference?

 

Comments

Popular posts from this blog

Top JavaScript courses helping you develop SPFx webPart with ReactJS and AngularJS

Enable the Microsoft Power BI report file type (.pbix) in SharePoint Search