torchdiffeq
torchdiffeq copied to clipboard
Stepping through backward pass of adjoint
Hello,
I would like to step through the code for the backward pass of the adjoint solver to gain a better understanding.
After adding the 'adjoint flag', I am able to step through the forward pass but not the backward one.
Would you have any idea why that is the case?
I am using pycharm. Thank you very much for your help
Unfortunately, I'm not sure how PyTorch's custom autograd functions works with a debugger either.. I use the cavemen method of prints and logs to debug. Maybe this can help: https://discuss.pytorch.org/t/custom-backward-breakpoint-doesnt-get-hit/6473/6
No problem, that is great. Thank you!