Brain-tumor-segmentation icon indicating copy to clipboard operation
Brain-tumor-segmentation copied to clipboard

Published 20 hours ago • verified publisher icon Issam28

Model does not converge

Open Gkakuka opened this issue 6 years ago • 4 comments

Hi Issam： thanks for you code！when I training the network，model does not converge. I try to reduce the learning rate and increase batch size，but it is not working. New to deep learning. Sorry for bothering you.

Gkakuka avatar Jan 14 '19 02:01 Gkakuka

why do you mean by model does not converge?

Issam28 avatar Jan 14 '19 10:01 Issam28

Sorry，I think that the loss of 0.6 is not small enough.

Gkakuka avatar Jan 14 '19 12:01 Gkakuka

Sorry，I think that the loss of 0.6 is not small enough.

Did you have a data augmentation? I can't reach the loss of 0.6, but 0.72

muween avatar Apr 06 '20 18:04 muween

Hi, I'm using train_test_split from sklearn for validation patches and labels. After passing the train and test values to fit_unet() function, I'm getting this error: 'ValueError: tf.function-decorated function tried to create variables on non-first call.' Please help me resolve this.

siddiqss avatar Aug 03 '20 11:08 siddiqss