🌲 TTool-AI is developed at the Laboratory for Timber Construction (director: Prof.Yves Weinand) with the support of the EPFL Center for Imaging, at EPFL, Lausanne, Switzerland. The project is part of the Augmented Carpentry Research.
🤖 TTool-AI automates the integration of new tools into AC (Augmented Carpentry), enhancing efficiency and simplifying the process. It is developed in Python and relies on the FastAPI framework. The project is containerized with Docker and Docker Compose. The Training Service is based on PyTorch. The project is developed and tested on Linux (Ubuntu 20.04) with NVIDIA GPUs.
🚀 For a quick hands-on start or more details, check out our Wiki.
- Go to the specified URL:
Visit the EPFL server at: http://128.178.91.106:16666/docs
- Follow the instructions:
Check out our Wiki for Users for more details.
-
Install Docker and Docker Compose:
Ensure you have Docker and Docker Compose installed on your system with NVIDIA Runtime support for the Training Service.
-
Environment Variables: TTool-AI relies on environment variables defined in a .env file. Make sure to set up this file as per the project's requirements.
Important: The .env file that is included in the repository is tailored for the EPFL server. You might have to modify the values for ${DATA_DIR}, ${VIDEO_DIR} and ${POSTGRES_DIR} to match your system's directory structure.
-
Clone the repository:
git clone git@github.com:ibois-epfl/TTool-ai.git
-
Run the project: Navigate to the project's root directory and run the following command:
cd TTool-ai/Run Docker Compose to build the project in the foreground:
docker compose up
-
Access the Service:
Once everything is up and running, you can access the FastAPI interface at:
-
If built on localhost: http://localhost:16666/docs
-
If built on a remote server: Use the appropriate IP address.
For more details, check out our Wiki.
For now the server can be accessed only by EPFL users via VPN. We are planning to make the server accessible from outside the EPFL domaine.