Generative AI with Python

2.11    Summary

In this chapter, you learned all about pretrained models, which are specialized, task-specific models.

This chapter began by looking back to a time before LLMs, when smaller models were developed for individual tasks using supervised learning techniques. These models excelled at specific tasks, were compact, and could often be run locally. In contrast, as you discovered, LLMs are built to handle a wide range of tasks using extensive, often unstructured data from various sources. While LLMs offer versatility, they come with higher costs and greater latency, making them less ideal for certain applications. We explored why it’s essential to choose between smaller, specific models and LLMs depending on your requirements and the resources ...