`rounding_mode` does not exist `.div` of PyTorch 1.3

[shahriar0651]

Tested with PyTorch 1.3, CUDA 10.0, and Python 3.3 with Conda.

class MaxPooling(Module):
.....
    def forward(self, input):
        output = SparseConvNetTensor()
        output.metadata = input.metadata
        output.spatial_size = (
            input.spatial_size - self.pool_size).div(self.pool_stride, rounding_mode='trunc') + 1
        assert ((output.spatial_size - 1) * self.pool_stride +
                self.pool_size == input.spatial_size).all()

But rounding_mode does not exist in PyTorch 1.3. Could you please update the following package requirements?

conda install pytorch torchvision cudatoolkit=10.0 -c pytorch # See https://pytorch.org/get-started/locally/