"AttributeError: 'tuple' object has no attribute 'shape'" in Keras Model

[Alper Kursat Uysal]

In the following sample code from the website " https://ai-leader.com/2020/05/09/using-lstm-with-1d-2d-and-3d-array/ ", I face with an error displaying "AttributeError: 'tuple' object has no attribute 'shape'". How can I get rid of that error ? Is there any size mismatch ?

 

# Using 3D array with LSTM

 

from keras import Model

from keras.layers import Input, Dense, Bidirectional

from keras.layers.recurrent import LSTM

import numpy as np

# define model for simple BI-LSTM + DNN based binary classifier

def define_model():

    input1 = Input(shape=(2,2)) #use row and column size as input size

    lstm1 = Bidirectional(LSTM(units=32))(input1)

    dnn_hidden_layer1 = Dense(3, activation='relu')(lstm1)

    dnn_output = Dense(1, activation='sigmoid')(dnn_hidden_layer1)

    model = Model(inputs=[input1],outputs=[dnn_output])

    # compile the model

    model.compile(loss='binary_crossentropy', optimizer='adam', metrics=['accuracy'])

    model.summary()

    return model

# Take a dummy 3D numpy array to train the model

data = np.array([[[0.1, 0.15], [0.2, 0.25]],

                 [[0.3, 0.35], [0.4, 0.45]],

                 [[0.5, 0.55], [0.6, 0.65]],

                 [[0.7, 0.75], [0.8, 0.85]],

                 [[0.9, 0.95], [1.0, 1.5]]])

Y = [1,1,1,0,0] #define binary class level for this model

print("data = ", data)

# NO NEED TO RESHAPE THE DATA as it is already in 3D format

# Call the model

model = define_model()

# Fit the model

model.fit([data],[np.array(Y)],epochs=4,batch_size=2,verbose=1)

# Take a test data to test the working of the model

test_data = np.array([[[0.2, 0.33],[0.2, 0.33]]])

# predict the sigmoid output [0,1] for the 'test_data'

pred = model.predict(test_data)

print("predicted sigmoid output => ",pred)

[Robert Kanasz]

Hi,

you can fix this problem by changing:

model.fit([data],[np.array(Y)],epochs=4,batch_size=2,verbose=1)

to

model.fit(data,np.array(Y),epochs=4,batch_size=2,verbose=1)

Regards

Robert