[Frontend][Pytorch] Add axis N when maxpool3d layout is (C,D,H,W)

[blackkker]

According to the pytorch doc torch.nn.MaxPool3d. The shape of input can be (N,C,D,H,W) or (C,D,H,W). While the MaxPool3DAttrs thinks the dimension should be 5. When importing a model contains MaxPool3d with shape (C,D,H,W)`,

tvm/include/tvm/relay/attrs/nn.h

Lines 985 to 996 in bcc7cde

	TVM_ATTR_FIELD(layout).set_default("NCDHW").describe(
	"Dimension ordering of input data. Can be 'NCDHW', 'NDHWC', etc."
	"'N', 'C', 'D', 'H', 'W' stands for batch, channel, depth, height, and width"
	"dimensions respectively. Pooling is applied on the 'D', 'H' and"
	"'W' dimensions.");
	TVM_ATTR_FIELD(out_layout)
	.set_default("")
	.describe(
	"Dimension ordering of output data. Can be 'NCDHW', 'NDHWC', etc."
	"'N', 'C', 'D', 'H', 'W' stands for batch, channel, depth, height, and width"
	"dimensions respectively. Pooling is applied on the 'D', 'H' and"
	"'W' dimensions.");

which could leads to out of bound error Check failed: (0 <= i && i < p->size_) is false: IndexError: indexing 4 on an array of size 4 when ii == 4 and dshape == 4.

tvm/src/relay/op/nn/pooling.cc

Lines 1211 to 1216 in bcc7cde

	if (dshape[ii].as<tir::AnyNode>()) {
	oshape[ii] = dshape[ii];
	} else {
	oshape[ii] =
	calculate_pool_dimension(dshape[ii], pad[i], param->pool_size[i], param->dilation[i],
	param->strides[i], param->ceil_mode);

Therefore, I simply add an expand_dims op on the frontend to avoid this situation.

[blackkker]

cc @masahi