Manage LLM Vector Databases and Long Term Memory in Flowise | AI Vector Tools Review | Part 2
2024 ж. 29 Қаң.
1 032 Рет қаралды
Learn how you can deploy Long Term Memory in Flowise and create continuous conversations with chat history. In Part 2 of this video, We'll look at three options you can use in Flowise to add long term to your AI chatflows as well as how to manage your token costs when it comes to chat history and input prompts.
Get Access to Chain Node Mastery (Flowise Mini-Course):
onyxstudiosinteractive.com/chain-node-mastery-mini-course/
Episode 5-6 Resource Page
On the resource page, you'll find all the links to the vector databases and tools mentioned in the review.
onyxstudiosinteractive.com/episode-5-6-llm-vector-long-term-memory-resources/
Need AI Consulting?
onyxstudiosinteractive.com/consulting/
Excellent content, subscribed, thanks
good stuff man, thanks🙌
😊 great video , thanks sir.
Excellent content. Congratulations on the 2 videos. Is there a way to use multiprompt in the example above, or do I need to create another type of flow? I would like to use multiprompt, but have the memory and vector retriever functionality.😄
It depends on what you're trying to do. For example, you could have a QnA chat with vector retrieval and memory like in the example linked to a series of prompts that are sequential. However the QnA chat would have to run first followed by the sequential prompts. In this way, the answers from the QnA chat would go to the prompts and serve as the info needed for the next question or task.
How can sessionId be made dynamic, meaning that each different user is assigned one?
You can set the sessionId variable in the API endpoint, provided that your memory node supports it. Zep has the sessionId variable along with Upstash and a few others. You can find out how to set the variable in the docs here: docs.flowiseai.com/integrations/langchain/memory/long-term-memory#ui-and-embedded-chat
@@OnyxStudiosInteractive thank you very much!
@@OnyxStudiosInteractive thank you for a direction to research! BTW, that would be great to have your video tutorial on production of multi-user chat bot too