Search results

torch.while_loop fails with AND of two conditions that succeed individually

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

module: guards

module: higher order operatorstorch.cond and similartorch.cond and similar

module: pt2-dispatcherPT2 dispatcher-related issues (e.g., aotdispatch, functionalization, faketensor, custom-op,PT2 dispatcher-related issues (e.g., aotdispatch, functionalization, faketensor, custom-op,

oncall: pt2

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.

#177517 In pytorch/pytorch;

· randolf-scholz opened on Mar 16, 2026

DDP init_sync=True does not sync buffers when broadcast_buffers=False

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

oncall: distributedAdd this issue/PR to distributed oncall triage queueAdd this issue/PR to distributed oncall triage queue

Status: Open.

#177514 In pytorch/pytorch;

· francois-rozet opened on Mar 16, 2026

[export] re-raising augmented exceptions resets traceback to except block

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

oncall: export

oncall: pt2

Status: Open.

#177510 In pytorch/pytorch;

· XVX-016 opened on Mar 16, 2026

[inductor][cpu] basic_gnn_sage fp32 single thread performance regression in 2025-12-15 nightly release

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

module: cpuCPU specific problem (e.g., perf, algorithm)CPU specific problem (e.g., perf, algorithm)

module: inductor

module: performanceIssues related to performance, either of kernel code or framework glueIssues related to performance, either of kernel code or framework glue

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

oncall: pt2

triage review

Status: Open.

#177509 In pytorch/pytorch;

· zxd1997066 opened on Mar 16, 2026

[inductor][cpu] torch_multimodal_clip AMP multiple thread performance regression in 2025-11-26 nightly release

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

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

Status: Open.

#177508 In pytorch/pytorch;

· zxd1997066 opened on Mar 16, 2026

torch.compiler.nested_compile_region: Repeated subgraph lifted constants break export, verification and ONNX serialization

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

module: onnxRelated to torch.onnxRelated to torch.onnx

module: pt2-dispatcherPT2 dispatcher-related issues (e.g., aotdispatch, functionalization, faketensor, custom-op,PT2 dispatcher-related issues (e.g., aotdispatch, functionalization, faketensor, custom-op,

oncall: export

oncall: pt2

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.

#177505 In pytorch/pytorch;

· vbaddi opened on Mar 16, 2026

[ROCm] test_add_layernorm: AssertionError: Expected 5 vec kernels, got 6

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

module: inductor

module: rocmAMD GPU support for PytorchAMD GPU support for Pytorch

module: testsIssues related to tests (not the torch.testing module)Issues related to tests (not the torch.testing module)

oncall: pt2

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.

#177501 In pytorch/pytorch;

· jsmolic opened on Mar 16, 2026

[ROCm] test_conv_stride_constraints: AssertionError: False is not true

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

module: convolutionProblems related to convolutions (THNN, THCUNN, CuDNN)Problems related to convolutions (THNN, THCUNN, CuDNN)

module: inductor

module: memory formatMemory format/layout related issues/changes (channels_last, nhwc)Memory format/layout related issues/changes (channels_last, nhwc)

module: rocmAMD GPU support for PytorchAMD GPU support for Pytorch

module: testsIssues related to tests (not the torch.testing module)Issues related to tests (not the torch.testing module)

oncall: pt2

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.

#177499 In pytorch/pytorch;

· jsmolic opened on Mar 16, 2026

-DUSE_ROCM=ON -DUSE_FLASH_ATTENTION=OFF -DUSE_MEM_EFF_ATTENTION=ON fails to build: use of undeclared identifier 'mha_fwd_aot'

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

module: buildBuild system issuesBuild system issues

module: rocmAMD GPU support for PytorchAMD GPU support for Pytorch

module: sdpaAll things related to torch.nn.functional.scaled_dot_product_attentiionAll things related to torch.nn.functional.scaled_dot_product_attentiion

triage review

Status: Open.

#177485 In pytorch/pytorch;

· AngryLoki opened on Mar 15, 2026

DISABLED test_comprehensive_masked_cumprod_xpu_float16 (__main__.TestInductorOpInfoXPU)

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

module: halfRelated to float16 half-precision floatsRelated to float16 half-precision floats

module: inductor

module: masked operatorsMasked operationsMasked operations

module: xpuIntel XPU related issuesIntel XPU related issues

oncall: pt2

skippedDenotes a (flaky) test currently skipped in CI.Denotes a (flaky) test currently skipped in CI.

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.

#177483 In pytorch/pytorch;

· etaf opened on Mar 15, 2026

scaled_dot_product_attention docs describe incorrect attn_mask shape

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

good first issue

module: docsRelated to our documentation, both in docs/ and docblocksRelated to our documentation, both in docs/ and docblocks

module: sdpaAll things related to torch.nn.functional.scaled_dot_product_attentiionAll things related to torch.nn.functional.scaled_dot_product_attentiion

smallWe think this is a small issue to fix. Consider knocking off high priority small issuesWe think this is a small issue to fix. Consider knocking off high priority small issues

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.

#177482 In pytorch/pytorch;

· josh-gleason opened on Mar 15, 2026

torch.compile silently succeeds on bmm/matmul with mixed dtypes (float16 @ float32) where eager raises RuntimeError

bot-triagedThis is a label only to be used by the auto triage botThis is a label only to be used by the auto triage bot

matrix multiplication

module: correctness (silent)issue that returns an incorrect result silentlyissue that returns an incorrect result silently

module: inductor

oncall: pt2

topic: fuzzer

triage review

Status: Open.

#177480 In pytorch/pytorch;

· Prorrrice opened on Mar 15, 2026