Nested Fourier Neural Operater for Darcy Flow

Core v0.4.0

This example demonstrates how to set up a data-driven model for a 2D Darcy flow using the Nested Fourier Neural Operator (FNO) architecture inside of Modulus.

Start with generating the dataset for training:

Copy
Copied!
            

python generate_nested_darcy.py

which will create the folder ./data with out_of_sample.npy, training_data.npy, validation_data.npy.

To train the model, run

Copy
Copied!
            

python train_nested_darcy.py +model=ref0 python train_nested_darcy.py +model=ref1

To evaluate the model use:

Copy
Copied!
            

python evaluate_nested_darcy.py

Previous Fourier Neural Operater for Darcy Flow
Next RNN for transient 2D Navier Stokes flow
© Copyright 2023, NVIDIA Modulus Team. Last updated on Jan 25, 2024.