torch.nn.MaxPool2d does not need "input" argument run but quantized function torch.nn.quantized.functional.max_pool2d must have it

[parth15041995]

quantized maxpool2d and adaptiveavgpool2d should not be defined in the quantizable version.