Custom Training Chatbot

This repository provides a framework for building a custom-trained chatbot using Llama and Streamlit. You can train the chatbot on your own data and use it for various tasks like question answering, information retrieval, or conversational interactions.

• Train on custom data: Train the chatbot on your own text documents or CSV files.
• Interactive interface: Streamlit interface allows you to upload files, query the chatbot, and manage data easily.
• Multiple data upload options: Upload data directly, via CSV, or through webhooks.
• Manage files: Delete unwanted data files directly from the interface.
• Configurable: Easy to modify the configuration file to adjust your chatbot's behavior.

Getting Started

Setting up the API

1. Download the API files:

git clone https://github.com/omkmorendha/Custom_Training_Chatbot/tree/api

2. Install the requirements:

pip install -r requirements.txt

3. Make an .env file and add OPENAI_API_KEY
4. Run the file by using

python main.py

Base URL:

http://127.0.0.1:5000

Authorisation Headers:

Authorisation : API_KEY Authorisation is done via checking if the SHA-256 hash of the API_KEY exists in the api_keys.txt file, currently we can use “abcd”, “efgh”, “ijkl” as the API key.

To use the API properly follow the documentation pdf in this repository

Setting up the Streamlit Web Application

1. Download the API files:

git clone https://github.com/omkmorendha/Custom_Training_Chatbot

2. Install the requirements:

pip install -r requirements.txt

3. Make an .env file and add OPENAI_API_KEY
4. Run the file by using

streamlit run app.py