Knowldge vs RAG #2205
Replies: 2 comments
-
so knowledge sources are integrated into an agent's memory system. They allow agents to access structured data from various formats like PDFs, text files, or web pages. When new knowledge is added, the agent's memory may need to be reset and re-initialized to incorporate this information. The RAG Tool dynamically retrieves information from diverse data sources during runtime, without requiring memory resets. It supports various data types, including PDFs, web pages, and YouTube videos, enabling agents to access up-to-date information as needed. |
Beta Was this translation helpful? Give feedback.
-
Hi, I've experimented with both options to return data from a csv. In both cases they are not returning all the expected rows. It should be 8 but 4 are returned instead. Any advise on how to get this fixed? Thanks, Alexandre |
Beta Was this translation helpful? Give feedback.
-
Hi,
I am wondering if someone can tell me what the difference is between CrewAI's Knowledge and creating a RAG function.
From the the docs I see that knowledge is based is memory based which means that if a new knowledge is added, the memory has to be reset, and agent is re-run again to include the latest knowledge.
Beta Was this translation helpful? Give feedback.
All reactions