Welcome to the Financial Bot project! This project demonstrates the setup of a retrieval-based question-answering (QA) chatbot that uses the langchain library for handling interactions and retrieval.
- Provides answers to user queries based on a database of financial information.
- Uses a custom prompt template to structure the QA interactions.
- Leverages langchain for language modeling and retrieval.
- Uses chainlit for managing the chatbot interactions.
- The
llama-2-7b-chat.ggmlv3.q4_0.bin
model can be run on systems with 8GB of RAM.
- Clone this repository:
git clone https://github.com/yourusername/financial-bot.git
- Install the required libraries:
pip install langchain chainlit
To install the required dependencies, you can use the provided requirements.txt
file. Simply run:
pip install -r requirements.txt
- Configure the
DB_faiss_path
variable to point to your local FAISS database. - Customize the prompt template in the
set_custom_prompt
function if needed. - Run the chatbot using
chainlit run app.py -w
.