Can not reproduce the results of NS

Thank you very much for your work, but I'm having trouble reproducing the NS equation. The accuracy is 0.0900, while the other equations are reproducible. I'm using the default parameters you used in your article, except for MLP_Ratio, which you didn't specify in your article. I used 1 and 2, respectively, but still can't achieve the desired accuracy. Could you provide the weights file?