TypeError: Cannot convert 0 to EagerTensor of dtype bool

[ShaniaShan]

Hi,
Thank you for sharing you wonderful work!

By using the print(), I find the values of variable Y and Y_v are zero matrix.

There are some errors when I run these codes as the below:
hist = unet.fit(X,tf.cast(Y,tf.float32),batch_size=32,epochs=100, validation_data=(X_v,tf.cast(Y_v,tf.float32)),verbose=1,callbacks= callback_list)

The errors are as the following:
TypeError: Cannot convert 0 to EagerTensor of dtype bool

 [[node EagerPyFunc (defined at <ipython-input-76-82292be68872>:11) ]] [Op:__inference_train_function_22126]

Errors may have originated from an input operation.
Input Source operations connected to node EagerPyFunc:
IteratorGetNext (defined at :17)

Function call stack:
train_function

[procton]

Same for me