LangFlow and Bubble - A Lethal Combo To Build a ChatPDF App!!!
Vložit
- čas přidán 26. 08. 2024
- Learn how to train an AI chatbot with your PDF. Configure and connect your LangChain app with this comprehensive tutorial video. Discover the step-by-step process of integrating LangFlow using the API connector in Bubble.io. Unlock the power of building your very own chat-PDF app from scratch with LangFlow's recent updates and complete revamp.
Exciting news! We're launching a comprehensive course that provides a step-by-step walkthrough of Bubble, LangChain, Flowise, and LangFlow. Join us to learn how to build multiple consumer-ready AI apps! Limited spots for early bird pricing builtbyou.com/
Link to Github Repo for LangFlow installation: github.com/men...
Link to Bubble app: bubble.io/page...
If you know of businesses or organizations looking to build a chat-with-document app, or any LangChain app, please feel free to let them know about Menlo Park Lab. We would love to help them out. Contact: info@menloparklab.com
Project requests: forms.office.c...
Link to Menlo Park Lab website: menloparklab.com/
For any help with implementation, join our discord: / discord
Link to LangFlow repo: github.com/log...
Who here is Team Langflow and who Team Flowise?
👀
This is gold! Thanks a lot for making this. ♥
Thanks for sharing the deploy template
Amazing amazing video. I am a subscriber now!!!
Great stuff Misbah! Is LangFlow included in your BuiltByYou course?
Thanks! Yes, Langflow is included.
I'd love to see this done with only local resources. I'm tied of having to pay for resources, when I have all the hardware I need to run everything locally.
Worth it...❤
hello, It would mean the world to me if you could answer my question beacouse I am not a tecnichal perssone and I don't know very well, I have a Limited Trial Plan on railway is this the problet that I can't deploy this?
so which do you prefer Langflow or Flowise?
Both are great and offer most of the functionality for use as a no-code lanchain backend. Also, teams supporting both products are very helpful. That said, I've personally used flowise a bit more and worked with the team closely. Just a personal preference.
Excellent
Thanks
Hi Sir, it would be nice to see how you can actually integrate your flowise bots in an existing website or intranet
Flowise offers an embed option that allows embedding the chat on any website, langflow doesn’t allow that yet, but most likely might release that in the future
@@menloparklab if i embedd it to a website as html code. Where is the Computing Power goin on? Does it still run over my pc, so i cannot turn it on off to use the bot?
can we connect redshift db to langflow? or any other db...
my pinecone always error how fix
Can you make video incorporate function calling (open ai) with bubble? Big thanks
On my list 😉
@@menloparklab Second this as well! Looking to do this functionality within Bubble straight from OpenAI API
@@menloparklab you are number 1
Hello, what's the risk of using langflow or flowise? If they stop their service for some reason will my app stop working?
Another great tutorial Misbah! Thank you! Question, doe your railway repo merges updates and new PR's? Or would that have to be done manually to update with new features?
Thanks, That would have to be done manually unless you have auto deploy option enabled for new commits/merges
But if you do that on Render/Railway, your previous flows will be deleted. Make sure to save them before updating.
Do you have a tutorial on how to setup LangFlow on Render? Great video by the way, really considering your course.
Langflow has a Dockerfile in the root folder so Render should instantly find it and set everything up.
@@ogabrielluiz thank you. The install line instructions, can I just rust run that? Would runtime be set to docker or python 3 then?
Nobody has done a tutorial on how to install Langflow on Render.
Great video! Thank you 😃 In your opinion which are the PRO and CONS of using Flowise/ LangFlow ? Is there one that it better for a specific task than the other?
PS: I've sent a Project Request for you throught the Form on the description! 🙏
Both are pretty good for retrieval applications. If you are looking to use more prebuilt agent tools, then langflow might have more options.
the problem i am facing is i uploaded a PDF of a company's annual report. then on asking questions like 'what was company's operating cash flow in 2022?', it says 'i dont know' when clearly the answer is mentioned in the report as 'cash flow from operating activities'...many such questions. e.g. capital expenditure and capex are the same but it won't find capex
I have come across some PDFs that are not easily parsable for text. My suggestion would be to check if text is fully extracted from the PDF before other retrieval steps.
I seen your this and Flowwise and Bubble videos. As per your prospective which one good Flowwise or Langflow?
Both offer unique solutions. My suggestions is to get the best of both. If you are building something using a Bubble UI, you can combine APIs from both.
Hi will your course have both langflow and flowise now? Or any updates? Also any of these you can use any TTS with? Whisper/11Labs/tortoise? Thanks
Yes, both langflow and flowise covered in the course. One could call the APIs to TTS services based on response from the LLM flow.
thanks for the video, I tried the same flow but why I send a message in the platform hatbox I got this error, Error: 'latin-1' codec can't encode characters in position 7-57: ordinal not in range(256). Any help?
Not sure what's causing that. Might need to try saving, then hitting the lightning button.
It's really awesome 🎉 ! Could you please also consider FlutterFlow as a no-code tool for other use cases with LangFlow? Thanks. 😊
Will try to go into FlutterFlow world soon.
Base on your experience which one is the bestt between LangFlow and Flowise? And in your video I saw Gpt-4 on LangFlow while in Flowise it doesn't go beyond Gp-3.5 I believe it's Davinci 3 or something like that. Can you explain why?
With the recent updates, LangFlow has really pushed the bar quite high. And since it’s based on Python langchain, it has a lot more integrations than Flowise.
@@menloparklab Yeah After your video I reached the same conclusion. Since your previous videos on Flowise I had been building some bots on it but I think I'll be switching to LangFlow
Both Flowise and LangFlow teams are quick to add new feature and are moving quite fast. It just that Python LangChain has a lot more integrations, and favors LangFlow as of now.
Hi Sir, how can i (manually) update flowise. As i understood it should auto update and redeploy on localhost:8080? it doesnt work for me even to run localhost:8080, im still at localhost:3000
Are you using the link shared to deploy on Railway?
@@menloparklab no, thank you👍👍
I this flow, is there a way to save the entire conversation after the interaction ends to review it?
Yes, one could save the chat with an ID for each user and the session, so that it can be reviewed later.
Can I retrieve any meta data around PDF with this solution? Let us say I want to write an essay on what have people written about langchain, I get multiple articles and I also get citations?
As long as you have embedded those written articles with their sources, then yes
@@menloparklab Jazak Allah Khair for your response, please do prepare a tutorial on that if you can 🙂.
I tried installing on Railway, and it has failed twice. It is on the trial account... Also, thank you, your videos are great.
I had to re-sync the fork of the repository, and it built.
repo deploy need paid railway account?
yup, that's true.
@@menloparklab I used the basic repo with render and it worked
Hi, I would be interested in your course. I have a few questions though. How can I contact you ? Also, I followed your video and tried to setup the community pdf loader to play with langflow. I tried for a while to setup the API in bubble, without success. I tried for ex :
{"inputs": {"input": ""},
"tweaks": [
{
"OpenAI-f7Rjz": {
"openai_api_key":""},
"PyPDFLoader-r2Kvp": {
"file_path":""}
Any chance you could help me on this ?
Thanks !
Best
Feel free to reach out to menloparklabai@gmail.com
Managed to solve? I bought the course and I have the same problem...