Skip to content

Latest commit

 

History

History
59 lines (42 loc) · 2.06 KB

README.md

File metadata and controls

59 lines (42 loc) · 2.06 KB

MOOZE

Overview

This repository contains a Python script that interacts with OpenAI's API to create and manage an AI assistant. The assistant is designed to respond in a specific style and help with questions related to a provided document. The script includes functionalities for managing files, vector stores, and threads to facilitate interactions with the AI assistant.

Prerequisites

  • Python 3.12
  • Python libraries:
    • dotenv library for environment variable management
    • openai library for interacting with OpenAI's API

Installation

  1. Clone the repository:

    git clone https://github.com/burakai/mooze.git
    cd mooze
  2. Install required packages:

    pip install -r requirements.txt
  3. .env file in the root directory of your project will be created automatically by the script in the new version:

    ... 
    OPENAI_API_KEY="your-openai-api-key"
    ...
    

Usage

  1. Initialize the Assistant:

    • The script starts by loading environment variables and setting up the OpenAI client.
    • It configures the assistant with predefined instructions and attempts to retrieve an existing assistant using assistant_id. If not found, it prompts to create a new assistant.
  2. Manage Threads:

    • The script tries to retrieve an existing thread using thread_id. If not found, it prompts to create a new thread.
  3. Manage Vector Stores:

    • The script attempts to retrieve an existing vector store using vector_store_id. If not found, the script exits.
  4. File Handling:

    • The script lists existing files and provides options to upload new files, continue with existing files, or start with no files.
    • Uploaded files are listed and used to create a vector store file.
  5. Interaction:

    • The script provides a continuous prompt for user input to interact with the assistant.
    • User inputs are sent to the assistant, and responses are retrieved and displayed.

Running the Script

To run the script, execute the following command in your terminal:

python main.py