How to Stream LangChainAI Abstractions and Responses using Streamlit Callback Handler and Chat UI

[@Avra_b]

Link to the blog post - medium.com/@avra42/stream-langchain-ai-abstractions-and-responses-in-your-web-app-langchain-tools-in-action-e37907779437

[@jasonferguson5929]

I am stumbling upon this video in March 2024 after trying many times to get my Streamlit / Langchain project to use outside tools... I'll be trying to jam this when I get home. Thank you so much for this one.

[@Avra_b]

Glad to hear ! Just a quick update, langchain package has been updated quite a time since over last couple of months. The code used might be deprecated already , however I’m certain they have something very similar ( streamlit callback handler )

[@phani3519]

do you have any tutorials on how to use vector database to query the document with streaming response?

[@Avra_b]

I do!

[@phani3519]

@@Avra_b hey man, I can say that I may not need those links anymore, way past that 😉

[@Avra_b]

@@phani3519 super! Sorry for the late response. Best wishes :)

[@adityaghongade3551]

Great tutorial !!! I have one question... If I am using "Human" tool of langchain which has a prompt which asks user to give some extra details for the agent to work properly, how are we going to leverage it to Streamlit. Could you make a video/ or give some suggestions for the same. It would be of great help. Thanks !

[@Avra_b]

I’m not sure langchain still have the same sdk

[@romellfudi]

Excelent tutorial!

[@Avra_b]

Thanks 😊 glad to hear 🤗

[@mikew2883]

Awesome video!👍

[@Avra_b]

Thank you ☺️

[@henkhbit5748]

Great video👍

[@Avra_b]

Thank you , appreciate

[@shivarajgudaganatti8573]

I want to stream thoughts etc. but without using streamlit as i have our own UI. Is it possible to do?

[@Avra_b]

No

[@ABHISHEKSHARMA-fo4tf]

This is amazing video where we can get lot of information . Can you make a video on how we connect streaming with llmchain and memory and show the streaming in webpage or html or streamlit?

[@Avra_b]

Thanks ! ☺️ the streaming callback handler must work. Irrespective of memory / chains.

[@ABHISHEKSHARMA-fo4tf]

the problem is when the response is come the after it is streaming , and it is just append the output so i make a receipe chatbot with vector database where i used the conversationchain , ones the chain is complete then it write in streaming , i want that it real time generating the text same as streaming so it is possible, beacause when it show after the response then it show time is double.can you help in this @@Avra_b

[@SarathSathyan-u2y]

This is great. But still there is no change in the waiting time know? only you are streaming the final text I guess

[@Avra_b]

Right ! It was way before ; things have changed now ;)

[@REALVIBESTV]

I sent you a message on Twitter but you haven't read the message

[@Avra_b]

Shoot here / or DM again please . Apologies . What’s your handle name ? Would love to connect

[@REALVIBESTV]

​@@Avra_bI've been researching Langchain and noticed that you have expertise in it. I'm currently working on an OpenAI voice chatbot in Python, but I lack knowledge in Langchain. I'm interested in integrating a conversation buffer memory into the bot so it can retain context during conversations. I don't know if you do work like that

[@REALVIBESTV]

I was asking if you take on a job like what am looking for

[@Avra_b]

Hi thanks for dropping your query . Sounds interesting. Unfortunately I’m not taking any freelance projects ATM . However , would you care to check out this video of mine - Build a PERSONAL CHATBOT with LangChainAI MEMORY with ChatGPT-3.5-Turbo API in PYTHON kzbin.info/www/bejne/mpKwf3p9ha-ee6c Also this -kzbin.info/www/bejne/mXnNnaObp6qGmMksi=V5xAUxu_a9_u6aY4 I’ve explained here the idea of using Memory in Chains . Try them out . You will find reference blog post as well . That might also add more context . Feel free to leave your comments in the underlying videos . Usually I make sure to reply to most of the comments , best wishes

[@eladiomendez8226]

Great stuff !

[@Avra_b]

Thanks ! I'm glad that you liked it :)