baby_agi_dspy

KO | EN

gitlite — search

#javascript #ai #go #golang #machine-learning #deep-learning #docker #java #windows #cli #awesome-list #pytorch #mcp #openai

(★ 15)

a version of baby agi using dspy and typed predictors

baby_agi_dspy Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Baby AGI System

The Baby AGI System is a Python-based automated goal-integrator (AGI) that dynamically generates, executes, and manages tasks to achieve a specified objective. Utilizing the power of the `dspy` library with an integration of OpenAI's GPT models, this system is designed to process complex objectives into actionable tasks, execute these tasks, and evaluate their outcomes until the overarching goal is reached.

## Features

- **Dynamic Task Management**: Automatically generates and manages a list of tasks based on the provided objective.
- **Task Execution and Monitoring**: Executes tasks and monitors their progress, updating their status as they are completed.
- **Objective Evaluation**: Continually evaluates the progress towards the objective, deciding when new tasks need to be created or when the objective has been achieved.

## Prerequisites

Before you can run the Baby AGI System, ensure you have the following installed:

- Python 3.7+
- `pydantic` for data validation and settings management.
- `python-dotenv` for environment variable management.
- The `dspy` library (ensure you have access to it or replace it with the actual library intended for use).
- An API key for OpenAI's GPT-3.5 Turbo model, which should be stored in an `.env` file as `OPENAI_API_KEY=your_api_key_here`.

## Installation

1. Clone this repository or download the source code.
2. Install the required dependencies:

```bash
    pip install pydantic python-dotenv
    # Install dspy library as per its installation instructions
    ```

3. Create a `.env` file in the root directory of your project and add your OpenAI API key:

```plaintext
    OPENAI_API_KEY=your_api_key_here
    ```

4. Load environment variables by running:

```bash
    python -m dotenv load_dotenv
    ```

## Usage

To run the Baby AGI System:

1. Navigate to the project directory in your terminal.
2. Execute the main script:

```bash
    python main.py
    ```

3. When prompted, enter the objective of your AGI. The system will then proceed to generate tasks, execute them, and provide updates on the progress towards achieving the specified objective.

## How It Works

- **Initiator Agent**: Generates an initial list of tasks required to start working towards the objective.
- **Task Agent**: Evaluates the current status of tasks and the objective to decide if new tasks are needed or if adjustments are required.
- **Execution Agent**: Takes individual tasks, executes them, and returns the results. It also determines if the current set of tasks and their results suffice to meet the objective.

The system operates in a loop, processing tasks generated by the Task Agent and executing them through the Execution Agent until the objective is achieved or no further actions are possible.

## Customization

The Baby AGI System is designed to be flexible. You can customize the `Task`, `TaskList`, and agent signatures (`InitiatorAgentSignature`, `TaskAgentSignature`, and `ExecutionAgentSignature`) in the provided Python script to suit your specific objectives and tasks.

baby_agi_dspy

baby_agi_dspy Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

baby_agi_dspy

baby_agi_dspy Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README