Operator Regression on Advection-Diffusion-Reaction (ADR) PDE (50 unseen operators)
0.0002RMSETransducer
Evaluation Results
| Method | Links | |
|---|---|---|
| TransducerTIME (S)=3.10e-3, GFLOPS=1.06e-1, Number of examples (n)=50, Optimization strategy=pre-trained2023.02 | 0.0002 | |
| FNOTIME (S)=1.72e2, GFLOPS=1.68e2, Number of examples (n)=50, Optimization strategy=optimized from scratch2023.02 | 0.0003 | |
| DeepONetTIME (S)=7.85e1, GFLOPS=1.54e2, Number of examples (n)=50, Optimization strategy=optimized from scratch2023.02 | 0.0202 |