0

I am incredibly confused as to the differences between the Tensorflow Class BinaryCrossentropy and the Function binary_crossentropy. The page where I see both on the documentation is here. The pages for the Class and the Function look nearly identical so if someone could explain it in a way that makes sense that would be fantastic.

Furthermore (but not really the main issue here) is when I pass the Function into my model.compile() it does not give me any errors, and what is even more confusing to me is that I can seemingly pass anything into model.compile() and running the line won't give any errors. Something like this:

model.compile(loss='gibberish') gives absolutely no errors, despite the gibberish loss function is defined nowhere. The documentation looks near identical so if someone could explain it in a way that makes sense that would be fantastic.

Thanks, Daniel

Daniel Sherman
  • 65
  • 4
  • let me reach you for 2nd item where you have put `model.compile(loss='gibberish')` , pls note that you will get the error when you try to fit the model `model.fit(x_train,y_train)` you will see that error will be thrown then. – simpleApp May 28 '21 at 01:48
  • also pls refer [binarycrossentropy-and-binary-crossentropy](https://stackoverflow.com/questions/59612914/difference-about-binarycrossentropy-and-binary-crossentropy-in-tf-keras-loss) – simpleApp May 28 '21 at 02:14

0 Answers0