jacklatrobe / YAGTA

YAGTA - Yet Another GPT Task Agent - Combines thinking from projects like BabyAGI and AutoGPT to create an autonomous task agent

YAGTA - Yet another autonomous task agent

YAGTA (Yet another autonomous task agent) is an experimental project that focuses on creating autonomous GPT (Generative Pre-trained Transformer) agents that can learn and improve over time. The goal of YAGTA is to develop an agent that can generate and execute tasks based on given objectives.

Getting Started

To get started with YAGTA, follow these steps:

Prerequisites

• Python 3.7 or higher
• OpenAI API key

Installation

1. Clone the YAGTA repository from GitHub:

git clone https://github.com/your-username/yagta.git

2. Navigate to the cloned repository:

cd yagta

3. Install the required dependencies:

pip install -r requirements.txt

Usage

1. Set up your OpenAI API key:

   • Go to the Azure portal and find your Azure OpenAI resource.
   • Retrieve your API key.

2. Run the main program loop:

   python main.py
   

   This will start the YAGTA agent and execute the tasks based on the given objectives.

Configuration

YAGTA uses the following components and configurations:

• LangChain: LangChain is a framework that simplifies the creation of applications using large language models (LLMs). It provides tools, components, and interfaces to manage interactions with language models.

• Vectorstore Memory: YAGTA uses VectorStoreRetrieverMemory to store and retrieve memories. It stores previous conversation snippets and uses an in-memory vectorstore for efficient retrieval.

• OpenAI LLM: YAGTA utilizes the OpenAI language model for generating responses and executing tasks. Make sure to set up your OpenAI API key before running YAGTA.

Contributing

Contributions to YAGTA are welcome! If you find any issues or have suggestions for improvements, please open an issue or submit a pull request on the GitHub repository.

License

YAGTA is released under the MIT License.