cb454682666fda678e5f2fabd4256d18bae6e50c,src/graph_transpiler/webdnn/frontend/keras/layers/convolutional.py,,_convert_conv2d,#Any#Any#,28

Before Change

    w = converter.convert_to_constant_variable(k_op.kernel, Order([Axis.KH, Axis.KW, Axis.C, Axis.N]))

    x, padding = convolution_handler_preprocess(x, ksize=k_op.kernel_size, padding=k_op.padding, dilation_rate=k_op.dilation_rate,
                                                data_format=k_op.data_format)
    y, = Convolution2D(None, ksize=k_op.kernel_size, stride=k_op.strides, padding=padding, dilation_rate=k_op.dilation_rate)(x, w)

After Change

@KerasConverter.register_handler("Conv2D")
def _convert_conv2d(converter: KerasConverter, k_op: "keras.layers.Conv2D"):
    x = converter.get_variable(converter.get_input_tensor(k_op)[0])
    check_data_format(x, k_op.data_format)

    w = converter.convert_to_constant_variable(k_op.kernel, Order([Axis.KH, Axis.KW, Axis.C, Axis.N]))

    paddings = (
        parse_padding(k_op.padding, k_op.kernel_size[0], k_op.dilation_rate[0]),
        parse_padding(k_op.padding, k_op.kernel_size[1], k_op.dilation_rate[1])
    )
    x, paddings = convert_odd_padding_to_concat(x, paddings=paddings)
    y, = Convolution2D(None, ksize=k_op.kernel_size, stride=k_op.strides, padding=paddings, dilation_rate=k_op.dilation_rate)(x, w)

    if k_op.use_bias:
        b = converter.convert_to_constant_variable(k_op.bias, OrderC)

Instances