Repository files navigation

The Art of Building Intelligent Applications with Langchain and Streamlit

This is a Streamlit app that allows users to chat with a PDF document using a conversational AI model. The app usesCohere for language modeling and question answering, andChroma for document indexing andLangchain for chaining all these together.

Use This Colab Notebook:Clickhereby@log-xp and@Nikhil-Paleti

For PDF Chatbot

https://chatwithpdf.streamlit.app/

https://huggingface.co/spaces/eswardivi/ChatwithPdf/

For Widgets (Streamlit Demo)

https://widgets.streamlit.app/

To use the app, follow these steps:

1. Upload a PDF document using the sidebar.
2. Type your message in the "You:" field and press "Send".
3. The AI model will generate a response based on the contents of the PDF document.
4. The response will be displayed in the chat window.

You can adjust the temperature of the AI model and the chunk size for splitting the document using the sliders in the sidebar.

Clone the project

  git clone https://github.com/EswarDivi/Anokha_Demo

Go to the project directory

cd Anokha_Demo

To use this app, you will need to create an account withCohere and get an API key. Once you have an API key, create a filesecrets.toml in the root directory of this project and add the following line:

cohere_apikey="<your_api_key>"

Install dependencies

    pip install -r requirements.txt

To deploy this project run

  streamlit run Talkwithpdf.py

To deploy this project on Streamlit Sharing, follow the steps below:

1. Create an account onStreamlit Sharing and connect it to your GitHub account.

2. Fork this repository to your GitHub account.

3. In the app secrets of your Streamlit Sharing dashboard, add a new secret namedcohere_apikey and set it to your Cohere API key.

4. Click onDeploy and wait for the deployment to finish.

5. Once the deployment is finished, you can access your app on the provided URL.

Note: Make sure your Cohere API key is kept secret and is not exposed to the public.

This app was created using the following libraries:

• Streamlit
• Cohere
• Chroma
• Langchain