- What is the recommended hardware for training using Modulus framework?
- What model architectures are in Modulus?
- What is the difference between Modulus Core and Symbolic?
- What can I do if I dont see a PDE in Modulus?
- What is the difference between the pip install and the container?
Please refer to the recommended hardware section: System Requirments
Nvidia Modulus is built on top of PyTorch and you can build and train any model architecture you want in Modulus. Modulus however has a catalog of models that have been packaged in a configurable form to make it easy to retrain with new data or certain config parameters. Examples include GNNs like MeshGraphNet or Neural Operators like FNO. Modulus samples have more models that illustrate how a specific approach with a specifc model architecture can be applied to a specific problem. These are reference starting points for users to get started.
You can find the list of built in model architectures here and here
Modulus core is the foundational module that provides the core algorithms, network architectures and utilities that cover a broad spectrum of Physics-ML approaches. Modulus Symbolic provides pythonic APIs, algorithms and utilities to be used with Modulus core, to explicitly physics inform the model training. This includes symbolic APIs for PDEs, domain sampling and PDE-based residuals. It also provides higher level abstraction to compose a training loop from specification of the geometry, PDEs and constraints like boundary conditions using simple symbolic APIs. So if you are familiar with PyTorch and want to train model from a dataset, you start with Modulus core and you import Modulus symbolic to bring in explicit domain knowledge. Please refer to the DeepONet example that illustrates the concept. If you are an engineer or domain expert accustomed to using numerical solvers, you can use Modulus Symbolic to define your problem at a higher level of abstraction. Please refer to the Lid Driven cavity that illustrates the concept.
Modulus Symbolic provides a well documeted example that walks you through how to define a custom PDE. Please see the source here to see the built-in PDE implementation as an additional reference for your own implementation.
There is no functional difference between the two. This is to simplify the ease of installing and setting up the Modulus environment. Please refer to the getting started guide on how to install using Pip or using the container.