Modulus vs modulus-sym & Post-processing (using the model after training)

[mitanshtrip]

Which modulus version should we use to run oour cases. I see that modulus-sym has the example cases and have been using that. For future cases we develop, should we use modulus or modulus-sym?

Secondly, once a model is trained, how do we use it for a live example? or deploy it for our purposes?

Thank you

[ngeneva]

Hi @mitanshtrip

Which modulus version should we use to run oour cases. I see that modulus-sym has the example cases and have been using that. For future cases we develop, should we use modulus or modulus-sym?

The goal of Modulus and Modulus-Sym are slightly different. Don’t think of these as versions, but rather different components / libraries of Modulus. Modulus (Core) is a more PyTorch developer orientated tool kit while Modulus-Sym provides an abstract framework with the symbolic loss representation. I.e. Are you interested in PINNs? Use Modulus-Sym. Are you a PyTorch developer looking at general Physics-ML, use Modulus (Core). Some discussion in our migration guide.

Secondly, once a model is trained, how do we use it for a live example? or deploy it for our purposes?

Please see the following thread:

Deploying/using a trained model | Post-Training

I’m currently trying to familiarize myself with the modulus platform (v22.09) and am a little confused as to how I am supposed to utilize the final models. Let’s say I’ve got a good model that I trained (foo.pth), and I now want to use it for generating predictions. The suggested modulus workflow documentation ends on the “training begins” step: https://docs.nvidia.com/deeplearning/modulus/modulus-v2209/user_guide/basics/modulus_overview.html How can I run the finalized model? E.g. is it the s…