Debugging and monitoring LangChain

By now, you’ve probably realized that even with the simpler chains, there is still a lot going on with LangChain under the hood. Often, when you’re constructing your prompts from different sources, you’ll need to see what the inputs and outputs of your language model are to understand and iterate on them.

So, it’s important to see what’s going on inside your application so that you can build, monitor, and debug your application. LangChain offers several different ways to achieve this.

Let’s explore these options to gain a clearer understanding of how they can enhance your application’s functionality. We’ll start with the simplest options and progress to the most comprehensive option: LangSmith.

Understanding tracing techniques

LangChain offers some simple ways to output the different steps of a chain.

If you’re working on the examples that are presented in this book using...