Modulus Launch is a PyTorch based deep-learning collection of training recipes and tools for creating physical surrogates. The goal of this repository is to provide a collection of deep learning training examples for different phenomena as starting points for academic and industrial applications. Additional information can be found in the Modulus documentation.
The recommended method for installing the latest version of Modulus Launch is using PyPi:
pip install nvidia-modulus.launchThe recommended Modulus docker image can be pulled from the NVIDIA Container Registry:
docker pull nvcr.io/nvidia/modulus/modulus:23.09For a local build of the Modulus Launch Python package from source use:
git clone git@github.com:NVIDIA/modulus-launch.git && cd modulus-launch
pip install --upgrade pip
pip install .To build Modulus Launch docker image:
docker build -t modulus-launch:deploy --target deploy -f Dockerfile .To build CI image:
docker build -t modulus-launch:ci --target ci -f Dockerfile .To build any of these images on top of the Modulus base image, you can
--build-arg BASE_CONTAINER=modulus:deploy to the above commands as shown below:
docker build --build-arg BASE_CONTAINER=modulus:deploy \
-t modulus-launch:deploy --target deploy -f Dockerfile .For guidance on making a contribution to Modulus, see the contributing guidelines
- Github Discussions: Discuss architectures, implementations, Physics-ML research, etc.
- GitHub Issues: Bug reports, feature requests, install issues, etc.
- Modulus Forum: The Modulus Forum hosts an audience of new to moderate level users and developers for general chat, online discussions, collaboration, etc.
Modulus Launch is provided under the Apache License 2.0, please see LICENSE.txt for full license text.