----------------------------------------------------------
MBEDDING_DIM = 70
cell = MinimalRNNCell(32)
activation_1 = keras.layers.advanced_activations.LeakyReLU(alpha=0.3)
activation_2 = keras.layers.advanced_activations.ELU(alpha=1.0)

model = Sequential()
model.add(Embedding(vocab_size, EMBEDDING_DIM, input_length = max_length))
model.add(GRU(units=64, dropout=0.2, recurrent_dropout = 0.2))


model.add(Dense(256, init='uniform'))
model.add(Activation(activation_1))
model.add(BatchNormalization())

model.add(Dropout(0.15))

model.add(Dense(64, init='uniform', activation='relu'))
model.add(Activation(activation_2))
model.add(BatchNormalization())

model.add(Dense(32, init='uniform', activation='relu'))
model.add(BatchNormalization())

model.add(Dropout(0.15))

model.add(Dense(1, activation='sigmoid'))

Model_Optimizer = optimizers.Adam(lr=0.0001, beta_1=0.9, beta_2=0.999, epsilon=None, decay=0.001, amsgrad=False)
model.compile(loss='sparse_categorical_crossentropy', optimizer=Model_Optimizer, metrics=['accuracy'])
model.compile(loss='binary_crossentropy', optimizer=Model_Optimizer, metrics=['accuracy'])


hist = model.fit(X_train_pad, Y_Train, batch_size=400, epochs=80,validation_split=0.2, verbose=1)