Denoising Pre-trained Operator Transformer for Navier-Stokes Equations

This example demonstrates how to set up a Denoising Pre-trained Operator Transformer for solving 2D Naiver-Stokes equation in PhysicsNeMo.

Note: This example runs on a single GPU. Multi-GPU training and pre-trained weights loading are in development.

Note: This example is a work in progress and will be updated in the next release.

Prerequisites

Install the required dependencies by running below:

pip install -r requirements.txt

To train the model, run

python train_dpot.py

training data will be generated on the fly.

Set train to False in config file to inference the model.

DPOT: Auto-Regressive Denoising Operator Transformer for Large-Scale PDE Pre-Training