From the course: Foundations of AI and Machine Learning for Java Developers

Unlock this course with a free trial

Join today to access over 24,500 courses taught by industry experts.

Prompts and completions

Prompts and completions

- [Instructor] How do you use a language model? How exactly do you program a large language model? We know that modern LLMs use a neural network, or a collection of neural networks, that contain the patterns of a very large amount of text. Certainly, it has learned a patterns of text from thousands and thousands of pages of text from the internet. And since it is a language model, the way that you program an LLM is through text. Let's say we are interested in having an LLM generate a text string based upon the text patterns it has memorized in its neural network. Well, what specific generated string? We have to start the language model with an initial string. This string is called a prompt. You want the large language model to create a new string based upon the patterns it knows given this initial string. So the prompt is guiding the behavior of the language model. You are setting it off in a certain direction. Based upon the patterns it has learned, you'll get a result that is very…

Contents