Chatbot Arena#

Demo

A live version of the application is hosted on Streamlit, try it out yourself using the link below: ChatBot Arena on Streamlit

This Streamlit application provides a user interface for interacting with Unify models through chat. It allows users to select models and providers, input text, and view the conversation history with AI assistants.

Developed with the software and tools below.

Overview#

This Streamlit application provides a user interface for interacting with Unify models through the chat. It allows users to select models and providers, input text, and view the conversation history with two AI assistants at a time. The app collects the data on the users’ assessment of the comparative models’ performance and provides an easy access to the global leaderboards which can be used as a complementary form of assessment of the performance of the models.

Motivation#

The challenge project “Chatbot arena” is based on this article.

Features#

  • Chat UI: Interactive chat interface to communicate with AI assistants.

  • Endpoint from Unify: Choose from a variety of models and providers.

  • Conversation History: View and track the conversation history with each model.

  • Clear History: Option to clear the conversation history for a fresh start.

  • Global Leaderboards: The votes are saved locally and globally.

How to use the app#

  1. Input Unify API Key: Enter your Unify API key in the provided text input box on the sidebar.

  2. Select endpoints : Choose the models and providers from the sidebar dropdown menus.

  3. Start Chatting: Type your message in the chat input box and press “Enter” or click the “Send” button.

  4. View Conversation History: The conversation history with the AI assistant for each model is displayed in separate containers.

  5. Clear History: You can clear the conversation history by clicking the “Clear History” button.

Getting Started#

System Requirements:

  • Python

  • streamlit

  • extra: look into the requirements.txt and requirements-test.txt files

Easy installation#

From source in order to use the attached Docker file.


Repository and Deployment#

Setup (without Docker)#

  1. Clone this repository:

    git clone https://github.com/samthakur587/LLM_playground
    
  2. change directory

    cd LLM_playground
    
  3. Install the required dependencies:

    pip install -r requirements.txt
    

Run the app#

streamlit run Chatbot_arena.py

Contributors#

Name

GitHub Profile

Samunder Singh

samthakur587

Kacper Kożdoń

Kacper-W-Kozdon

---