Modifying manually pytorch state_dict file and retrain

Asked Apr 29 '19 at 09:57

Active May 28 '22 at 13:15

Viewed 1,138 times

I save and load pytorch state_dict file and then evaluate and retrain the model. That works.

torch.save(net.state_dict(), path)
net.load_state_dict(torch.load(path))

However, when I modify the state_dict file (manually changing the values) after loading it and evaluating it, I receive the error:

RuntimeError: a leaf Variable that requires grad has been used in an in-place operation.

How to safely modify the state_dict and then retrain it without the error?

asked Apr 29 '19 at 09:57

cerebrou

please give a minimal code example of the way you modify the `state_dict`. – Shai Apr 29 '19 at 12:46
[This](https://stackoverflow.com/a/49316176/1714410) might answer your question – Shai Apr 29 '19 at 12:48

0 Answers0