RAGLight supports three main search modes:
- Standard RAG: the base retrieval enhances the generation process by vectorizing the query to retrieve similar documents and using them as context to generate answers
- Agentic RAG: Implemented through AgenticRAGPipeline, it adds the function of intelligent body, supports multi-step reasoning and dynamic adjustment of search strategy.
- RAT (Retrieval Augmented Thinking): Implemented through the RATPipeline, the addition of a reflection step (reflection parameter) significantly improves the logic and accuracy of the answer.
The choice of mode depends on the complexity of the application scenario. Standard RAGs can be used for simple questions and answers, Agentic RAGs are used when multi-step reasoning is required, and RAT modes are used when accuracy is required.
This answer comes from the articleRAGLight: Lightweight Retrieval Augmentation Generation Python LibraryThe
May not be reproduced without permission:AI productivity tools " What are the search modes supported by RAGLight? What are the characteristics of each?
English 
简体中文 
日本語 
Deutsch 
Português do Brasil