Chat with your PDF Chatbot: All OPEN SOURCE (Runs on CPU)

In this tutorial video, I will guide you through the process of creating your own PDF Chatbot using open-source technologies. With the power of conversational AI, you'll be able to enhance your interaction with PDF documents using just your CPU.

To get started, we'll be using the Langchain framework for text preprocessing and handling chains. This framework provides robust capabilities for processing text, ensuring that your chatbot can understand and respond effectively to your queries.

Next, we'll leverage Sentence Transformers, a powerful embedding model, to enable intelligent understanding and retrieval of information. By incorporating this model, our chatbot will be able to grasp the context of your queries and provide accurate and relevant responses.

For efficient storage and retrieval of vectorized information, we'll utilize the Chroma DB, a high-performance vector database. This will ensure quick access to the relevant information and enhance the overall responsiveness of your chatbot.

To give your chatbot its language capabilities, we'll make use of the Lamini LM 738M, a large language model known for its natural language processing capabilities. With this model as the backbone, your chatbot will be able to provide insightful and contextually relevant responses to your queries.

To create a user-friendly interface, we'll employ Streamlit, a web framework that allows for easy interaction with your PDF Chatbot. With its intuitive design and interactive features, you'll be able to seamlessly engage with your chatbot and explore its capabilities.

By following this tutorial, you'll gain the knowledge and skills to build your own PDF Chatbot using open-source technologies. So let's dive in and unlock the potential of conversational AI for your PDF documents!

#langchain #chatbot #python

[harshkadam]

Thank you Buddy, for creating this video ! 😊

[aimanrajpoot]

I'm new to AI and have created a Flask app to build a REST API. Although my model works correctly and responds accurately, I'm struggling to deploy the app. I attempted to deploy it on Heroku, but unfortunately, it didn't succeed. My ultimate goal is to use this API in Android apps by making API calls to interact with my model.

[jas.sin]

Thanks for the well explained video. I just have a question. I want to scale this app for multiple PDFs stored on a drive or some location. Would need your inputs on the same and would be helpful for me.

[VenkatesanVenkat-fdhg]

Hi, Thanks for the valuable videos and recently watched llm sagemaker deployment. My doubts are

1. How you handled error on lambda related to IAM policy whether it is specifically for accessing sagemaker endpoints here...

2. For getting api response, whether we do not need any flask or fastapi implementation.

Can you guide on this. waiting for ur responses and videos....

[harshkadam]

is this open source model is powerful to write like blog, poem anything means like we can give input field for user to put there own text like if its blog writer tool, if we pass text in the text field and enter, then this text pass to the our prompt we have created like prompt template like : Write blog on [this topic] with Seo . [this topic] is like user input bascilally . that is just example there is lots of potential to make any type of tools with prompting

[harshkadam]

i also heard like there is bard unofficial API means like its like cookies which we can able to use as API for free, I know its not like official, its just cookies using, I think there is also potential to build some tools with it

[divyeshvora]

Hi there,

Could you kindly provide the necessary installation list to streamline our work on the code? Since everyone has a different Python version and machine configuration, having this list will be very helpful.

I also have a question: You mentioned that your code works on CPU. Does your deep learning command not require a GPU?

I apologize for my basic question; I'm new to Python. Thank you in advance for your help.

Thanks

[gnanaprakashm]

Hi, thanks for the video. I am currently doing a project that uses trains open source models like google flan t5, falcon etc using some scraped data (in csv / txt) format. How do I acheive that ?

[naveenpandey]

Thank you for this tutorial.

I have tried Document Summarizer where i uploaded large PDF and it got crashed. I think same thing will happen with this application if i upload large pdf😅

[brussesserunkuuma]

hi there thanks for your great work this is the ultimate starter place for practical hands on no time to waste learnings though i have only one request if you could please reference the prerequisite under the vids or the about would be really help full tho thank you very much

[John-jxho]

Thanks for these great informative videos. I have a question, how to create a chatbot that answers based upon given pdf only (not user uploaded but hard-coded)

[imranurrehman]

very nice and informative video. can you please guide how we can make chat bot for a specific website. handling all conversations and information with clients and visa versa . thankuou

[ricardofield]

Hi, thanks for all the video. I usually deploy chatbots in spanish, teh accuracy for this models gets lower using multilanguage models. But great info what you are posting, congrats.
Are you thinking on posting something using RAG for PDF retrieval? thanks

[anup..]

Getting below error, while running streamlit - Could you please help.

User Warning: The installed version of bits and bytes was compiled without GPU support. 8-bit optimizers, 8-bit multiplication, and GPU quantization are unavailable. warn ("The installed version of bits and bytes was compiled without GPU support. ")

[akhilpadmanaban]

Watching all the AI vides is kinda confusing because everyone uses Open AI, and their things...I donno how can we create things if they allow just a bit of token length to be accessible and the rest being paid..I know thats a business strategy, so instead we can go for our own implementations by learning just how this model works no?

[harshkadam]

Sir, I am also python developer, but I am not able to build complex Ai apps still , can you suggest what should I specifically learn to get into Ai, is there any courses ? Like how should you learn this things, can you share some resources to learn ai ml in python

[codexperience]

hi, may i know if this is free and open source or there's part that need to get a free api key but will expire after a certain period?

[CharanjitSingh-lbqq]

Great explanation very helpful wondering can we use faiss instead of chromadb if yes please guide

[AkashDesai-efmk]

Can I use this without any limits or restrictions? plzz reply

[berfinpala]

Does anybosy got the error "MarshallComponentException: ('Could not convert component args to JSON', TypeError('Object of type Document is not JSON serializable'))" on the line message(history["generated"][i], key=str(i)) ?? Couldn't find a solution so far