Colorfu1

Confusion about SpecialSpmmFunction

1

In official documents, there is a notice 'Each function object is meant to be used only once (in the forward pass).' in subclass of torch.autograd.Function. In model SpGraphAttentionLayer, you have...

I noticed that SpecialSpmmFunction is the subclass of torch.autograd.Function and there is only one object in class SpGraphAttentionLayer. `class SpGraphAttentionLayer(nn.Module): def __init__(self, in_features, out_features, dropout, alpha, concat=True): super(SpGraphAttentionLayer, self).__init__() self.in_features...

I use scatter_max with 1800w+ points occur illegal CUDA memory. It is Ok with 1500w+ points or with cpu operator. torch 2.3.1+cu118 torch-scatter 2.1.2 usage: **new_feat, argmax = torch_scatter.scatter_max(feat, unq_inv,...