In this project, we are going implement a multiple document question and answer We can import many PDFs and chat using OpenAI or HuggingFace API The frontend framework is Streamlit, a powerful tool to fast prototype and implement
In order to achieve a great result, we will use the best available technologies and methods through Deep Learning models
- 📚 'Streamlit official site' - link
- 📚 'Langchain official site' - link
- 📚 'OpenAI official site' - link
- 📚 'HuggingFace official site' - link
-
Clone the repository to your local machine
-
Create a python .venv environment and activate it
-
Install the requirements -> pip install requirements.txt
-
Generate your access tokens on OpenAI and HuggingFace account
-
Create a file called ".env" and insert the bellow info:
5.1 OPENAI_API_KEY='your_key'
5.2 HUGGINGFACEHUB_API_TOKEN='your_key' -
Execute streamlit at terminal -> streamlit run app.py
-
Just use it on browser
-
Load the files on the sidebar
-
Click on a 'process' button to process the documents
-
Start chatting
- Test other APIs and Embeddings
- Deploy on AWS