demo.py

import tensorflow
from tensorflow.keras.layers import Lambda, Dense, Input
from tensorflow.keras.models import Model
from tensorflow.keras import backend as K

def loss_fn(args):
        return K.constant(1, dtype = 'float32')# creating model
inputs = Input(shape = (784,))
dense1 = Dense(512, activation = 'relu')(inputs)
dense2 = Dense(128, activation = 'relu')(dense1)
dense3 = Dense(32, activation = 'relu')(dense2)

# create classification output
classification_output = Dense(10, activation = 'softmax')(dense3)

outputs = Lambda(loss_fn, name = 'loss', output_shape = (1,))(classification_output)
model = Model(inputs = inputs, outputs = outputs)

model.compile(tensorflow.keras.optimizers.Adam(learning_rate = 0.01), loss = lambda y_true, y_pred: y_pred)
~