LLM Apps: Evaluation course

This repository contains the code for the LLM Apps: Evaluation course.

Learn to build reliable evaluation pipelines for LLM applications by combining programmatic checks with LLM-based judges. Develop techniques for automated evaluation, from writing effective criteria to aligning automated scores with human judgment.

For more LLM, MLOps and W&B platform courses visit AI Academy.

Setup

1. Create a new conda environment using the provided requirements.txt:

conda create --name eval-course --file requirements.txt