ResUnet-a icon indicating copy to clipboard operation
ResUnet-a copied to clipboard

Published 20 hours ago • verified publisher icon Akhilesh64

How to continue to train?

Open GulerEnes opened this issue 3 years ago • 3 comments

Hi, I am trying to use your codes on Google Colab. But sometimes Colab suddenly closes itself. Is there a way to continue training?

I didn't see any flag in main.py about this.

GulerEnes avatar May 04 '22 18:05 GulerEnes

Hi @GulerEnes this is an issue with colab. It shuts down if no activity is detected for sometime. There is no problem with the code.

Akhilesh64 avatar May 04 '22 19:05 Akhilesh64

Hi @GulerEnes this is an issue with colab. It shuts down if no activity is detected for sometime. There is no problem with the code.

Yes, I know the codes are working well. Each 10 epoch a new model.h5 file saving. I just want to continue the training from the last saved model.h5 for any reason. How can I do this?

GulerEnes avatar May 04 '22 19:05 GulerEnes

The scripts don't yet support resuming from checkpoint but I'll try to implement this if I get time or you can do these changes yourselves because this seems pretty straightforward.

Akhilesh64 avatar May 10 '22 06:05 Akhilesh64