具有不确定性量化的函数

[ayiiiii1105]

您好，
我是一个初学者，当我复制[学习具有不确定性量化的函数](https://github.com/lululxvi/deepxde/blob/master/examples/function/func_uncertainty.py] 中包含的完整代码运行时发生了

    net = dde.nn.FNN(
        layer_size,
    ...<3 lines>...
        dropout_rate=0.01,
    )
TypeError: FNN.__init__() got an unexpected keyword argument 'dropout_rate'

报错，询问ai后得到当前 DeepXDE 版本的 FNN 不支持 dropout_rate 参数的结果，但我已经尝试更新为最新的deepxde版本，请问为什么会报错？我又该如何解决这个问题。

[ayiiiii1105]

Hello,
I'm a beginner. When I copied the complete code from Learning Functions with Uncertainty Quantization and ran it, I encountered an error. After consulting AI, I learned that the current version of DeepXDE's FNN does not support the dropout_rate parameter. However, I have already tried updating to the latest DeepXDE version. Why is this error occurring, and how can I resolve it?

 File "D:\PythonProject\求解不确定量化方程.py", line 24, in <module>
    net = dde.nn.FNN(
        layer_size,
    ...<3 lines>...
        dropout_rate=0.01,
    )
TypeError: FNN.__init__() got an unexpected keyword argument 'dropout_rate'

[ayiiiii1105]

I accidentally clicked "close issue" instead.Additionally, I'd like to ask if PyTorch, as a backend, cannot perform multi-fidelity computation? I encountered the error "module 'deepxde.nn.pytorch' has no attribute 'MfNN'" when learning Multi-fidelity learning from the formulas. How can I resolve this?

[ayiiiii1105]

@lululxvi Mr. Lu, could you please reopen this issue for me? Thank you！

[lululxvi]

Try TensorFlow backend.