Essential Guide to LLMOps

4

Developing Models via LLMOps

In this chapter, we’ll cover how to develop a large language models (LLM) while ensuring that LLMOps best practices are followed. This ensures that the developed LLM can be effectively reviewed and eventually deployed to production. The information in this chapter exemplifies a real-world use case for developing a performant LLM through LLMOps. We will be looking at the following topics:

• Creating features
• Storing features
• Retrieving features
• Selecting foundation models
• Fine-tuning models
• Tuning hyperparameters
• Automating model development

Creating features

Features are derived from processed data typically used as input to models. Creating features in a distributed computing environment is fundamental to developing ...