BasicRAG

Simple single-call RAG: retrieve once, build prompt, generate once.

Overview

Field	Value
Type	Generation
Algorithm	Retrieve + Generate
Modality	Text

How It Works

1. Retrieve top-k documents using configured retrieval pipeline
2. Build context from retrieved documents
3. Generate answer using LLM with prompt template

Configuration

_target_: autorag_research.pipelines.generation.basic_rag.BasicRAGPipelineConfig
name: basic_rag
retrieval_pipeline_name: bm25
llm: gpt-4o-mini
prompt_template: |
  Context:
  {context}

  Question: {query}

  Answer:
top_k: 5
batch_size: 100

Options

Option	Type	Default	Description
name	str	required	Unique pipeline instance name
retrieval_pipeline_name	str	required	Name of retrieval pipeline to use
llm	str or BaseLLM	required	LLM instance or config name
prompt_template	str	default	Template with {context} and {query}
top_k	int	10	Documents to retrieve
batch_size	int	100	Queries per batch

Prompt Template Variables

Variable	Description
{context}	Retrieved document contents
{query}	Original query

When to Use

Good for:

• Simple Q&A tasks
• Baseline RAG implementation
• Quick prototyping

Consider advanced pipelines for:

• Multi-hop reasoning
• Iterative retrieval
• Complex answer synthesis

Citation

@article{lewis2020retrieval,
  title={Retrieval-augmented generation for knowledge-intensive nlp tasks},
  author={Lewis, Patrick and Perez, Ethan and Piktus, Aleksandra and Petroni, Fabio and Karpukhin, Vladimir and Goyal, Naman and K{\"u}ttler, Heinrich and Lewis, Mike and Yih, Wen-tau and Rockt{\"a}schel, Tim and others},
  journal={Advances in neural information processing systems},
  volume={33},
  pages={9459--9474},
  year={2020}
}