Skip to content

Latest commit

 

History

History
56 lines (39 loc) · 2.27 KB

README.md

File metadata and controls

56 lines (39 loc) · 2.27 KB
Raw

NitroGPT Navigator: GPT-3.5 Based AI ChatBot

NitroGPT Navigator is an advanced chatbot that leverages cutting-edge technologies, including OpenAI's GPT-3.5, LangChain, LLMs and Pinecone Vector Database. It's designed to provide accurate and context-aware answers to any questions related to the content on routerprotocol.com.

Features

  • Powered by GPT-3.5: NitroGPT Navigator utilizes the power of OpenAI's latest language model to generate human-like responses.
  • LangChain Integration: LangChain enhances linguistic capabilities, enabling a more nuanced understanding of user queries.
  • Vector Embeddings: Using a Large Language Model from HuggingFace, all-MiniLM-L6-V2, to convert Text to Numeric form for generating Vector Embeddings.
  • Pinecone Vector Database: Leveraging Pinecone allows for efficient vector similarity searches, improving the accuracy of information retrieval.

How It Works

NitroGPT Navigator operates by processing user queries through a multi-layered approach:

  1. User Input: Users submit questions related to routerprotocol.com.
  2. LangChain Processing: LangChain parses and understands the user's query, extracting key information.
  3. Vector Search with Pinecone: Pinecone performs vector searches to identify relevant content.
  4. GPT-3.5 Response Generation: OpenAI's GPT-3.5 generates context-aware responses based on the extracted information.

Installation

To run NitroGPT Navigator locally, follow these steps:

  1. Clone the repository:

    git clone https://github.com/ShivankK26/NitroGPT-Navigator.git
cd NitroGPT Navigator

  2. Install dependencies:

    pip install -r requirements.txt

  3. Set up API keys:

    • Obtain API keys for OpenAI, Hugging Face, and Pinecone.
    • Add these keys to the corresponding configuration files.

  4. Run the application:

    streamlit run app.py

Usage

Once the application is running, users can interact with RouterBot by asking questions related to routerprotocol.com. The chatbot will provide informative and context-aware responses.

Example:

What is Router Nitro?

Contributing

Contributions are welcome! If you have ideas for improvements, bug fixes, or new features, feel free to open an issue or submit a pull request.