What you'll learn

Working with large language models (LLMs) does not always yield the best results. In this course, Context Evaluation and Safety, you’ll learn to build metrics-based tests and interpret the differences between hallucination and faithfulness. First, you’ll explore using Python and evaluation libraries like DeepSeek to test your model by evaluating output against ground truths. Next, you’ll discover how to set up a scoring system to evaluate the ratio of the supported claims as a test of fairness. Finally, you’ll learn how to build automated processes to periodically test your model. When you’re finished with this course, you’ll have the skills and knowledge of context evaluation and safety validations needed to build your own LLM-as-a-judge.