RAGAS: A New Framework for Evaluating RAG

Use Oxen AI 🐂 https://oxen.ai/ Oxen AI makes versioning your datasets as easy as versioning your code! Even is millions of unstructured images, the tool quickly handles any type of data so you can build cutting-edge AI. -- Paper 📜 https://arxiv.org/abs/2309.15217 Dataset 🔢 https://www.oxen.ai/ox/Arxiv-Dive-RAG Links + Notes 📝 https://www.oxen.ai/blog/arxiv-dives Join Arxiv Dives 🤿 https://oxen.ai/community Discord 🗿 https://discord.com/invite/s3tBEn7Ptg -- Chapters 0:00 Intro to the arXiv Dive 1:05 What is RAGAS? 2:15 Quick Recap on What RAG is 4:29 Intro to the arXiv Dive Bot 7:38 Why Evaluating RAG is Hard 8:26 Intro to the Example Dataset 13:19 Enter RAGAS 15:20 The First Criteria: Faithfulness 24:23 The Second Criteria: Answer Relevance 29:27 The WikiEval Dataset 32:15 The Third Criteria: Context Relevance 40:45 Conclusion