I tried computing gradients for implementation of Physics informed deep learning but I kept getting runtime error:
RuntimeError: One of the differentiated Tensors does not require grad.
I Have tried all possible solutions that I cam across online but not successfully resolved it. Below is the code for computing my first and second derivative.
def fwd_gradients(Y, x):
dummy = torch.ones_like(Y)
G = torch.autograd.grad(Y, x, grad_outputs=dummy, create_graph=True)[0]
Y_x = torch.autograd.grad(G, x, grad_outputs=torch.ones_like(G), create_graph=True)[0]
return Y_x