An overview of OpenAI Playground
Today, OpenAI offers a set of pre-trained, ready-to-use models that can be consumed by the general public. This has two important implications:
- Powerful foundation models can be consumed without the need for long and expensive training
- It’s not necessary to be a data scientist or an ML engineer to manipulate those models
Users can test OpenAI models in two ways:
- OpenAI Playground, a friendly user interface where you can interact with models without the need to write any code.
- Via REST API with a pro-code approach. In fact, all OpenAI models – and, generally speaking, LLMs – comes as pre-built components that can be consumed by applications via endpoints and keys.
In this chapter, we will explore the first approach, while we will see how to incorporate OpenAI models’ APIs in Chapter 11.
Before jumping into the Playground, let’s first have an overview of OpenAI’s models families.
OpenAI Models Families
Over...
