def get_siamese_model(input_shape):

    left_input = Input(shape=input_shape)

    right_input = Input(shape=input_shape)

    model = Sequential()

    model.add(Embedding(input_dim=max_token, output_dim=512, input_shape=input_shape))

    model.add(Conv2D(64, (10,10), name="conv1", activation='relu',

                   kernel_initializer=initialize_weights, kernel_regularizer=l2(2e-4)))

    model.add(MaxPooling2D())

    model.add(Conv2D(128, (7,7), name="conv2", activation='relu', kernel_initializer=initialize_weights,

                     bias_initializer=initialize_bias, kernel_regularizer=l2(2e-4)))

    model.add(MaxPooling2D())

    model.add(Conv2D(128, (4,4), name="conv3", activation='relu', kernel_initializer=initialize_weights,

                     bias_initializer=initialize_bias, kernel_regularizer=l2(2e-4)))

    model.add(Flatten())

    model.add(Dense(4096, activation='sigmoid',

                   kernel_regularizer=l2(1e-3),

                   kernel_initializer=initialize_weights,bias_initializer=initialize_bias))

    encoded_l = model(left_input)

    encoded_r = model(right_input)

    L1_layer = Lambda(lambda tensors:K.abs(tensors[0] - tensors[1]))

    L1_distance = L1_layer([encoded_l, encoded_r])

    prediction = Dense(1,activation='sigmoid',bias_initializer=initialize_bias)(L1_distance)

    siamese_net = Model(inputs=[left_input,right_input],outputs=prediction)

    return siamese_net