863a9dcee691e850e79d97a16abb977e24324fa1,tests/convolution_test.py,ConvTest,test_convlayer_3d_bn_reg_prelu_shape,#ConvTest#,104

Before Change

    def test_convlayer_3d_bn_reg_prelu_shape(self):
        x_3d = self.get_3d_input()
        conv_reg = ConvolutionalLayer(
            10, 3, 2,
            w_regularizer=regularizers.l2_regularizer(0.5),
            with_bias=False,
            with_bn=True,
            acti_func="prelu")
        conv_reg_out = conv_reg(x_3d, is_training=True)
        print(conv_reg)
        with self.test_session() as sess:
            sess.run(tf.global_variables_initializer())
            out_3d = sess.run(conv_reg_out)
            self.assertAllClose((2, 8, 8, 8, 10), out_3d.shape)

After Change

                                           is_training=True)

    def test_convlayer_3d_bn_reg_prelu_shape(self):
        input_param = {"n_output_chns": 10,
                       "kernel_size": [5, 1, 2],
                       "stride": [1, 1, 2],
                       "with_bias": False,
                       "with_bn": True,
                       "acti_func": "prelu",
                       "w_regularizer": regularizers.l2_regularizer(0.5),
                       "b_regularizer": regularizers.l2_regularizer(0.5)}
        self._test_conv_layer_output_shape(rank=3,
                                           param_dict=input_param,
                                           output_shape=(2, 16, 16, 8, 10),
                                           is_training=True)

    def test_convlayer_3d_relu_shape(self):
        input_param = {"n_output_chns": 10,
                       "kernel_size": [5, 1, 2],

Instances