Issues
- 1
ARM64 torch_cluster.radius bug
#217 opened - 1
Missing `omp.h` error
#216 opened - 3
Err: source code compile
#215 opened - 0
Simplify build wheels github workflow pipeline
#212 opened - 2
- 2
Unable to pip install torch-cluster
#209 opened - 5
- 4
setup.py depends on `torch`
#204 opened - 4
Still there: Incompatibility with bfloat16
#203 opened - 4
- 3
RuntimeError: Not compiled with CUDA support,help please.my torch is 2.1.0 and my cuda is 12.0.
#201 opened - 2
warning when work with torch.compile
#200 opened - 2
can k >100 in knn?
#199 opened - 2
- 1
Binaries for PPC (ppc64le) architecture.
#197 opened - 2
How to run fps algorithm in editable mode?
#196 opened - 8
- 2
No such operator torch_cluster::fps_ptr_list
#193 opened - 2
documentation
#192 opened - 3
CUDA 12.1 binaries
#190 opened - 2
- 3
Support Python 3.11 for conda packages
#187 opened - 11
- 10
Poetry / PEP 517 build support
#185 opened - 4
Install torch-cluster for C++
#184 opened - 1
Documentation error?
#182 opened - 5
- 2
Support for AMD GPUs (ROCm) via HIP?
#178 opened - 4
- 7
- 4
How to sample k points in fps
#174 opened - 2
Variable K in KNN
#173 opened - 3
- 5
about radius and radius_graph
#171 opened - 10
- 2
- 2
- 11
Issue running with multi gpu
#165 opened - 8
INTERNAL ASSERT FAILED for knn for k > 100
#164 opened - 6
Illegal memory access with GPU implementation of nearest() if batch_y includes an empty instance
#163 opened - 3
- 3
- 1
Error Compiling with Python 2.0
#159 opened - 1
Inconsistency in graph creation
#158 opened - 1
- 3
'nearest' slows down after one run
#155 opened - 9
gpu execution error
#148 opened - 1
Speed issues
#147 opened - 2
cannot install on windows11
#146 opened - 3
Incompatibility with bfloat16
#145 opened