neuraloperator
neuraloperator copied to clipboard
neuraloperator
Learning in infinite dimension with neural operators.
In the updated FNO2d model a 'padding' was introduced. According to a comment in the code this should help with non-periodic inputs. I think this padding may break the ability...
Benchmark
Hi Zongyi Li, Incredible work, very interesting! I am studying this subject in my PhD. But I am having some difficulty in obtaining the same results as figure 2 of...
Due to the feature of cudnn and cuda, the random-seeds-fixing part of previous code is not complete, so reruning the same program still returns different (and in some way uncontrollable)...
Hi, is burgers_v1000_t200_r1024_N2048.mat the burgers' equation with viscosity=1000 or viscosity=1/1000? Could you please help me out here? Thanks!
Hi, Can I use the fourier2d.py for data containing 1 spatial dimension and 1 time dimension? Thanks Soumyadipta
Hello o/, I am reproducing your great work! But failed to find the dataset specified in fourier_3d.py file as below: TRAIN_PATH = 'data/ns_data_V100_N1000_T50_1.mat' TEST_PATH = 'data/ns_data_V100_N1000_T50_2.mat' And below is your...
hello, may I interrupt you for a moment recently I want to reproduce the application of FNO to Bayesian inverse problems (FOURIER NEURAL OPERATOR FOR PARAMETRIC PARTIAL DIFFERENTIAL EQUATIONS section...
Hello Zongyi ! I like your project / paper / code and I use it in the company I work for (It gives better result that unet, is faster and...
https://github.com/zongyi-li/fourier_neural_operator/blob/90bbd4c9ffa84b91ec2daa86660148861989adec/fourier_3d.py#L53 What you can do is using FFTshift: ``` def roll_n(X, axis, n): f_idx = tuple(slice(None, None, None) if i != axis else slice(0, n, None) for i in range(X.dim()))...
Metadata
Owner
Metadata
Learning in infinite dimension with neural operators.