Search results

NCCL-2.28.3 build locally is unusable on H100

actionable

high priority

module: buildBuild system issuesBuild system issues

module: ciRelated to continuous integrationRelated to continuous integration

module: ncclProblems related to nccl supportProblems related to nccl support

module: regressionIt used to work, and now it doesn'tIt used to work, and now it doesn't

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

#164402 In pytorch/pytorch;

· malfet opened on Oct 1, 2025

Runtime failure when running torch.compile() & using GCC 11.5.0 on Neoverse V1

actionable

module: armRelated to ARM architectures builds of PyTorch. Includes Apple M1Related to ARM architectures builds of PyTorch. Includes Apple M1

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

module: third_party

module: vectorizationRelated to SIMD vectorization, e.g., Vec256Related to SIMD vectorization, e.g., Vec256

oncall: cpu inductorCPU Inductor issues for Intel team to triageCPU Inductor issues for Intel team to triage

oncall: pt2

Status: Open.

#162422 In pytorch/pytorch;

· Rohanjames1997 opened on Sep 8, 2025

Building cuda kernels with debug information causes some cuda kernel launches to fail due to capacity constraints

module: buildBuild system issuesBuild system issues

module: cudaRelated to torch.cuda, and CUDA support in generalRelated to torch.cuda, and CUDA support in general

module: debug-buildRelated to building and testing PyTorch in debug modeRelated to building and testing PyTorch in debug mode

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

#160225 In pytorch/pytorch;

· galv opened on Aug 8, 2025

Fully remove the protoc dependency in core

module: buildBuild system issuesBuild system issues

module: onnxRelated to torch.onnxRelated to torch.onnx

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

#159156 In pytorch/pytorch;

· albanD opened on Jul 25, 2025

RuntimeError: optimized_plan_ INTERNAL ASSERT FAILED at "/opt/conda/envs/bld/conda-bld/libtorch_1741644046984/work/torch/csrc/jit/runtime/profiling_graph_executor_impl.cpp":725, please report a bug to PyTorch.

module: third_party

needs reproductionEnsure you have actionable steps to reproduce the issue. Someone else needs to confirm the repro.Ensure you have actionable steps to reproduce the issue. Someone else needs to confirm the repro.

oncall: jitAdd this issue/PR to JIT oncall triage queueAdd this issue/PR to JIT oncall triage queue

Status: Open.

#158701 In pytorch/pytorch;

· qiqicliff opened on Jul 19, 2025

[CUDA 12.9 | sm_120] Build fails on RowwiseScaledMM.cu with PTXAS error (PyTorch 2.8.0a0)

module: buildBuild system issuesBuild system issues

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

module: cudaRelated to torch.cuda, and CUDA support in generalRelated to torch.cuda, and CUDA support in general

module: third_party

needs reproductionEnsure you have actionable steps to reproduce the issue. Someone else needs to confirm the repro.Ensure you have actionable steps to reproduce the issue. Someone else needs to confirm the repro.

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

#156613 In pytorch/pytorch;

· celsowm opened on Jun 23, 2025

Importing xgboost before torch + openmp causes seg fault

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

module: macosMac OS related issuesMac OS related issues

module: openmpRelated to OpenMP (omp) support in PyTorchRelated to OpenMP (omp) support in PyTorch

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

#155201 In pytorch/pytorch;

· ikrommyd opened on Jun 5, 2025

MPS topk failure for 5D tensor or above

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

module: mpsRelated to Apple Metal Performance Shaders frameworkRelated to Apple Metal Performance Shaders framework

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

Bug

#154890 In pytorch/pytorch;

· manuelcandales opened on Jun 2, 2025

MPS batch_norm mixed dtype failure

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

module: mpsRelated to Apple Metal Performance Shaders frameworkRelated to Apple Metal Performance Shaders framework

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

Bug

#154887 In pytorch/pytorch;

· manuelcandales opened on Jun 2, 2025

MPS max_pool2d_with_indices failure: destination values and indices length mismatch along axis

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

module: mpsRelated to Apple Metal Performance Shaders frameworkRelated to Apple Metal Performance Shaders framework

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

Bug

#154882 In pytorch/pytorch;

· manuelcandales opened on Jun 2, 2025

MPS cumsum failure for 5D tensor or above

high priority

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

module: mpsRelated to Apple Metal Performance Shaders frameworkRelated to Apple Metal Performance Shaders framework

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

Bug

#154881 In pytorch/pytorch;

· manuelcandales opened on Jun 2, 2025

[RFC] Removing ideep git submodule dependency from PyTorch & add oneDNN git submodule

module: armRelated to ARM architectures builds of PyTorch. Includes Apple M1Related to ARM architectures builds of PyTorch. Includes Apple M1

module: intelSpecific to x86 architectureSpecific to x86 architecture

module: third_party

triagedThis issue has been looked at a team member, and triaged and prioritized into an appropriate moduleThis issue has been looked at a team member, and triaged and prioritized into an appropriate module

Status: Open.

#154444 In pytorch/pytorch;

· sanchitintel opened on May 27, 2025