Darcy 2D l2.backward() error

[saki606]

Hoping someone can clear up the error below.
Tryting to run MGKN_general_darcy2d.py and encounter the following error during Training (see below).
Seems like a variable is being overwritten (in_place?) somewhere . . . can't figure out exactly where.
Tried variations of adding variable.clone() in the NN forward feed process to see if it helps, but nothing so far.

---

preprocessing finished, time used: 257.48266062999755
3 [0, 400, 500, 525]
Traceback (most recent call last):
File "MGKN_general_darcy2d.py", line 280, in
l2.backward()
File "/home//projects/dir_env/env_pytorch/lib/python3.8/site-packages/torch/_tensor.py", line 396, in backward
torch.autograd.backward(self, gradient, retain_graph, create_graph, inputs=inputs)
File "/home//projects/dir_env/env_pytorch/lib/python3.8/site-packages/torch/autograd/init.py", line 173, in backward
Variable._execution_engine.run_backward( # Calls into the C++ engine to run the backward pass
RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [525, 64]], which is output 0 of ReluBackward0, is at version 1; expected version 0 instead. Hint: enable anomaly detection to find the operation that failed to compute its gradient, with torch.autograd.set_detect_anomaly(True).

adding torch.autograd.set_detect_anomaly(True) produces the following message:

---

RuntimeError: one of the variables needed for gradient computation has been modified by an inplace operation: [torch.cuda.FloatTensor [525, 64]], which is output 0 of ReluBackward0, is at version 1; expected version 0 instead. Hint: the backtrace further above shows the operation that failed to compute its gradient. The variable in question was changed in there or anywhere later. Good luck!