README.md

---
license: mit
tags:
- desktop-application
- ai-tools
- dataset-management
- lora-training
- windows
---

# ADAM — AI Development and Automation Manager

ADAM is a local Windows desktop hub for organizing AI image-development workflows. It helps you prepare and review datasets, connect your own training tools, plan runs with explicit approval, monitor jobs, and keep a history of your own assets.

## What is included

- ADAM source code and the built-in tool registry.
- A clean, empty external-tool registry.
- No model weights, LoRAs, checkpoints, datasets, generated images, job history, personal paths, API keys, or local settings.

## Requirements

- Windows 10/11
- Python 3.10 or later
- Optional: NVIDIA GPU for compatible training workflows
- Optional: Ollama for local chat assistance

## Install and run

```powershell
git clone https://huggingface.co/SyntheticMDProductions/AI_Development_Automation_Manager
cd AI_Development_Automation_Manager
python -m pip install -r requirements.txt
python main.py
```

On first launch, ADAM creates your personal `config/settings.json` automatically. In **Settings → Tool folders**, connect the local projects, datasets, and models that you own and want ADAM to manage. ADAM does not bundle or download model weights for you.

## Notes for users

- Training and collection plans require approval before ADAM starts them.
- You are responsible for the licenses, permissions, and rights for any datasets, models, and third-party tools you connect.
- This repository is a downloadable desktop application. It is not a hosted Hugging Face Space or an inference model.

## Development

```powershell
python -m pytest -q
```

## License

ADAM is released under the [MIT License](LICENSE).