Closed Sehmimul closed 11 months ago
I can do the first task, saving a loss for DCG. Right now, we do not pass the testloader into DCG, so computing accuracy requires few more steps.
I did the task 1. Check issue #16 for the same
If this is implemented, we can mark it as completed. @Sehmimul
Issue addressed in https://github.com/vishnu-m77/DiffAPTOSC/issues/36
DCG is a classifier which is pre-trained. We need to find loss curve of pre-trained dcg and also compute metrics like accuracy.
The tasks are: