Search results

AArch64 unit test failure test_transpose_mxn_16_16_bf16_fp16 on Neoverse-V1

Status: Open.

#170877�?In pytorch/pytorch;

· robert-hardwick opened on Dec 19, 2025

Type promotion logic for tensor operations(tensor & tensor combination ops) sucks,its'nt supporitng for shell dtypes,its justifiable, but even though the tensors are of the same dtype-> not returning the output tensor with the same dtype,(throwing error now)

Status: Open.

#170873�?In pytorch/pytorch;

· gouthamk1926 opened on Dec 19, 2025

Feedback about torch.addcdiv

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

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.

#170869�?In pytorch/pytorch;

· mcw-anasuya opened on Dec 19, 2025

Operator benchmark: option to measure GPU execution time only (less CPU noise)

Status: Open.

#170867�?In pytorch/pytorch;

· apakbin opened on Dec 19, 2025

[RFC][Inductor] Support the judgment of training or inference scenarios for custom post grad fx pass

oncall: pt2

Status: Open.

#170866�?In pytorch/pytorch;

· cdzhan opened on Dec 19, 2025

[BUG][Dynamo] Host segfault/coredump caused by passing a device tensor pointer to a generated CPU kernel

oncall: pt2

Status: Open.

#170863�?In pytorch/pytorch;

· peaceorwell opened on Dec 19, 2025

DISABLED test_template_bad_epilogue_fusion_use_async_compile_False (__main__.TestEpilogueFusionStaticAnalysis)

module: xpuIntel XPU related issuesIntel XPU related issues

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.

#170854�?In pytorch/pytorch;

· etaf opened on Dec 19, 2025

DISABLED test_view_as_complex_xpu (__main__.TestViewOpsXPU)

module: xpuIntel XPU related issuesIntel XPU related issues

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.

#170853�?In pytorch/pytorch;

· etaf opened on Dec 19, 2025

Add forward_by_train and forward_by_eval convenience methods to nn.Module

module: nnRelated to torch.nnRelated to torch.nn

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.

#170851�?In pytorch/pytorch;

· msdsm opened on Dec 19, 2025

Varlen attention API with custom varlen <code>attn_mask</code>

Status: Open.

#170839�?In pytorch/pytorch;

· vadimkantorov opened on Dec 19, 2025

MPS backend - inconsistent results for batched inference on BERT/RoBERTa

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

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

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.

#170837�?In pytorch/pytorch;

· jsrozner opened on Dec 19, 2025

Generated torch.Library autograd functions seem not have <code>setup_context</code> which breaks <code>torch.func.grad</code> usage

module: custom-operatorscustom operators, custom ops, custom-operators, custom-opscustom operators, custom ops, custom-operators, custom-ops

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.

#170834�?In pytorch/pytorch;

· vadimkantorov opened on Dec 19, 2025