Keep consistency in Classes ConvBNActivation

[zhiqwang]

It seems that the class ConvBNReLU has been renamed to ConvBNActivation, but it still retains in the super(ConvBNReLU, self).__init__() method, it will make users a little confused, so I changed it to ConvBNActivation in this PR.

vision/torchvision/models/mobilenetv2.py

Lines 32 to 55 in 54d88f7

	class ConvBNActivation(nn.Sequential):
	def __init__(
	self,
	in_planes: int,
	out_planes: int,
	kernel_size: int = 3,
	stride: int = 1,
	groups: int = 1,
	norm_layer: Optional[Callable[..., nn.Module]] = None,
	activation_layer: Optional[Callable[..., nn.Module]] = None,
	dilation: int = 1,
	) -> None:
	padding = (kernel_size - 1) // 2 * dilation
	if norm_layer is None:
	norm_layer = nn.BatchNorm2d
	if activation_layer is None:
	activation_layer = nn.ReLU6
	super(ConvBNReLU, self).__init__(
	nn.Conv2d(in_planes, out_planes, kernel_size, stride, padding, dilation=dilation, groups=groups,
	bias=False),
	norm_layer(out_planes),
	activation_layer(inplace=True)
	)
	self.out_channels = out_planes

[datumbox]

Good spot! I have indeed missed this when I changed it.

I left one optional recommendation for your consideration, let me know.