GOLDEN_SET.md
Gold-standard reference outputs for evaluation
14 documents availableCategories
Recent GOLDEN_SET.md Documents
View allGolden Dataset Guidelines
Defines a structured format for a golden dataset used to benchmark, regression test, and compare retrieval-augmented generation systems.
Project Memory
Documents a complete RAG system for Paris cultural events with architecture, requirements, and implementation history across 15 phases.
AWS Certified Generative AI Developer – Professional (AIP-C01)
Provides study notes and prep guidance for the AWS Certified Generative AI Developer, Professional (AIP-C01) exam, including exam style, prerequisites, and service summaries.
Understanding the Sources of Uncertainty - and Why Our Evals are Biased
Catalogues nine sources of error in AI evaluations, bias and noise, with concrete examples and statistical intuition for each.
IR-Copilot — Incident Response AI Assistant
Serves as a reference implementation for an agentic RAG system with multi-tenant chat, document ingestion, hybrid search, and subagent delegation.
📈 Trading RAG Mentor
Answers trading questions using your own video transcripts via a RAG pipeline with ChromaDB, LangChain, and a choice of Groq or Ollama LLMs.
Evaluation Scripts
Provides two Python scripts: one interactively builds a golden test set of queries, the other evaluates retrieval quality using IR metrics.
Understanding the Sources of Uncertainty - and Why Our Evals are Biased
Catalogues 10 sources of error in AI evaluations, bias, noise, and their compounding effects, with worked examples and statistical intuition.