23Aug02 | How to talk to ChatGPT | For Devs

All concepts are discussed in this short deeplearning.ai course.

Principles

• Principle I: Write clear and specific instructions
• Principle II: Give the model time to “think”

How to reduce model hallucinations?

1. Find relevant info/quotes from the text
2. Ask it to use those quotes to answer questions

Iterative Prompt Development

Iteratively analyze and refine your prompts:

1. Try something
2. Analyze where the result doesn’t give what you want
3. Clarify instructions, give it more time to think
4. Refine prompts with a batch of examples

Example:

In practice, you’d end up with a prompt like this only after multiple iterations. No way around it.

Summarizing

Several techniques to get the most out of GPT’s summarizing capabilities:

Inference

In just a few minutes you can make multiple systems for making inferences from text that would’ve taken many days or weeks for a skilled ML developer!

Note: Output of LLMs is pretty inconsistent, so to be able to use its output in a prod env, we’d need to force a JSON output, and then parse it.

Transforming

Also there’s translation!

Expanding

Temperature