Phoenix Metrics for RAG Evaluation: Key Insights and Options
How many metrics does Phoenix provide for RAG evaluation?? Are there any metrics called Response_Evaluation?
Right now we only have relevance for evaluating document nodes but you can build your own document evaluator and Phoenix will calculate all the retrieval metrics for you based on that. https://docs.arize.com/phoenix/llm-evals/building-your-own-evals For response evaluations - hallucinations and QA relevance are the main two that take into account context faithfulness. Let us know if there are metrics you're looking for Sangram!