RAG (Retrieval Augmented Generation)
technique technique
Technique to enhance LLMs with external knowledge
Supported languages
Concepts
embeddingsvector-searchchunks
Pros and Cons
Ventajas
- + Up-to-date knowledge
- + Reduces hallucinations
- + Private data without fine-tuning
- + Cites sources
Desventajas
- - Additional complexity
- - Quality depends on retrieval
- - Higher latency
Casos de Uso
- Enterprise chatbots
- Document Q&A
- Assistants with specific knowledge