when I used the sigmoid and binary cross entropy ,I have found that the train_acc is very high about 98% at the begining and I didn't use fine-tune. I use img/255. normalized the train img and i set the label value 0 and 1……so what is wrong with it？…
Donate. I desperately need donations to survive due to my health
Get paid by answering surveys Click here
Click here to donate
Remote/Work from Home jobs
Comments
Post a Comment