xfail('sparse.sampled_addmm', ''), # RuntimeError: Sparse CSR tensors do not have strides xfail('sparse.mm', 'reduce'), # RuntimeError: Sparse CSR tensors do not have strides # Non-contiguous Bugs # @@ -567,6 +
It looks like the adjacency matrix cannot be multiplied by itself because it is a sparse tensor as opposed to when the edge weights aren't given and are given a one for all edges. Changing the tensor multiplication function used here should fix the issue, but I still haven't managed to ...
To associate your repository with the sparse-matrix topic, visit your repo's landing page and select "manage topics." Learn more Footer © 2025 GitHub, Inc. Footer navigation Terms Privacy Security Status Docs Contact Manage cookies Do not share my personal information ...
Support sparse tensorsShigekiKarita/grain#8 weiyangfb mentioned thison Jul 17, 2018 [WIP] add mul_ / mul(S, D) and sparse_mul_ / sparse_mul(S, D) in ATen#9481 40remainingitems Load more ezyang commentedon Jul 23, 2020 ezyang
const at::sparse_csr::SparseCsrTensor& C) { #if defined(CUDA_VERSION) && CUDA_VERSION < 11000 #if (!defined(USE_ROCM)) && (defined(CUDA_VERSION) && CUDA_VERSION < 11000) TORCH_CHECK( false, "Calling addmm with sparse GPU tensors requires compiling ", "PyTorch with CUDA 11+. ", ...