I’m facing problems loading a custom LSTM Module. The model works as expected during the training phase, but when I save the model and load it back, I get very different and very inaccurate results.
I have also verified the weights of all the layers of the model before and after loading, and they are exactly similar, but the results are significantly inaccurate after loading.
I have tried all the methods described in the official doc (https://pytorch.org/tutorials/beginner/saving_loading_models.html)
None of the methods seem to work.
Can someone please help me regarding the issue.