Part 08 · Evaluating LLM Apps

LLM-as-a-Judge

Using a model to grade a model: writing judge prompts, pairwise comparisons, the biases judges have, and checking the judge against human labels.

⏱ 1-2 hours · Station 64 of 92