guided-diffusion icon indicating copy to clipboard operation
guided-diffusion copied to clipboard

How can I judge the training of classifier have been diverged?

Open blakery-star opened this issue 2 years ago • 2 comments

I've tried to train a classifier for a new dataset,how can I judge when to stop the training?

blakery-star avatar Oct 27 '22 17:10 blakery-star

@blakery-star Have you figure out the solution? And I also would like to ask what's your classifier train and val accuracy. Thanks in advance!

ptoyip avatar May 04 '23 11:05 ptoyip

@ptoyip Hi, have you tried to train the classifier? The classifier is trained on noisy images, so I assume that it would have quite low top-1 and top-5 accuracies.

AlexofNTU avatar Aug 01 '23 12:08 AlexofNTU