PDE Operator Learning on NS-PwC
0.1EGFNO
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FNOTraining Regime=Trained from Scratch2024.05 | 0.1 | 1 | |
| CNOTraining Regime=Trained from Scratch2024.05 | 0.37 | 1.5 | |
| scOTTraining Regime=Trained from Scratch2024.05 | 0.54 | 2 | |
| MPP-BTraining Regime=Pretrained2024.05 | 0.74 | 2.3 | |
| CNO-FMTraining Regime=Pretrained2024.05 | 1.66 | 3.3 | |
| DPOTTraining Strategy=From Scratch, Model Size=M, Number of parameters=120 M2024.05 | 17 | 6.1 | |
| DPOTTraining Strategy=From Scratch, Model Size=L, Number of parameters=509 M2024.05 | 23.3 | 10.2 | |
| DPOTTraining Strategy=Finetuned, Model Size=L, Number of parameters=509 M2024.05 | 39.7 | 12 | |
| DPOTTraining Strategy=Finetuned, Model Size=M, Number of parameters=120 M2024.05 | 44.8 | 12.5 | |
| POSEIDON-LTraining Regime=Pretrained2024.05 | 89.06 | 24.7 |