pytorch_sparse icon indicating copy to clipboard operation
pytorch_sparse copied to clipboard

Published 20 hours ago • verified publisher icon rusty1s

GCNConv (or any) fails for input of type SparseTensor.

Open shash29-dev opened this issue 3 years ago • 3 comments
trafficstars

ASSUME a Pytorch geometric layer - Layer(x, edge-index) <--- sucess Sped= SparseTensor.from_edge_index(edge_index) Layer(x,Sped) <--- Fails with Assertion

Why is this issue?

shash29-dev avatar Sep 20 '22 09:09 shash29-dev

What is the error you encounter? :) This should definitely work.

rusty1s avatar Sep 20 '22 09:09 rusty1s

Hi, I tried the same on linux, it runs, with sparse tensor. While on windows, Sparse tensor support fails with device side assert for input of type Sparsetensor.

shash29-dev avatar Sep 22 '22 12:09 shash29-dev

Do you have a minimal example to reproduce? Does the same happen on CPU? At least in our CI, at least Windows CPU works fine.

rusty1s avatar Sep 23 '22 05:09 rusty1s

This issue had no activity for 6 months. It will be closed in 2 weeks unless there is some new activity. Is this issue already resolved?

github-actions[bot] avatar Mar 23 '23 01:03 github-actions[bot]