4cf94b4f157424e41be9cd4f45a9d0b3cf1bbfe5,dlpy/blocks.py,DenseNetBlock,compile,#DenseNetBlock#Any#Any#,382

Before Change

        concat_num = 1
        input_layer = src_layer
        for layer in self.layers:
            if layer.config["type"].lower() in ("convo", "convolution"):
                layer.name = "D{}C{}".format(block_num, conv_num)
                conv_num += 1
                layer.src_layers = [input_layer]
            elif layer.config["type"].lower() == "batchnorm":
                layer.name = "D{}B{}".format(block_num, bn_num)
                bn_num += 1
                layer.src_layers = [input_layer]
            elif layer.config["type"].lower() == "concat":
                layer.name = "D{}Concat{}".format(block_num, concat_num)
                concat_num += 1
                layer.src_layers = [input_layer, src_layer]
                src_layer = layer

After Change

            A dictionary of keyword-arguments

        """
        if block_num is None:
            block_num = self.get_number_of_instances()

        options = []
        conv_num = 1
        bn_num = 1
        concat_num = 1
        input_layer = src_layer
        for layer in self.layers:
            if layer.type == "convo":
                layer.name = "D{}C{}".format(block_num, conv_num)
                conv_num += 1
                layer.src_layers = [input_layer]
            elif layer.type == "batchnorm":
                layer.name = "D{}B{}".format(block_num, bn_num)
                bn_num += 1
                layer.src_layers = [input_layer]
            elif layer.type == "concat":
                layer.name = "D{}Concat{}".format(block_num, concat_num)
                concat_num += 1
                layer.src_layers = [input_layer, src_layer]
                src_layer = layer

Instances