Search results

<code>fork()</code> is no longer the default start method in Python 3.14

Status: Open.

#169252 In pytorch/pytorch;

· njzjz opened on Nov 30, 2025

Large discrepancy between CPU and GPU norms for identical float32 tensor

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

Status: Open.

#169237 In pytorch/pytorch;

· vinith2 opened on Nov 29, 2025

[MPS] nn.ConvTranspose{1,2,3}d ignore validation check for output_padding &lt; stride (Silent Failure)

actionable

module: error checkingBugs related to incorrect/lacking error checkingBugs related to incorrect/lacking error checking

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.

#169236 In pytorch/pytorch;

· lingebeng opened on Nov 29, 2025

[MPS] nn.MaxUnpool{1,2,3}d fails to check shape consistency between input and indices (Silent Failure)

actionable

module: error checkingBugs related to incorrect/lacking error checkingBugs related to incorrect/lacking error checking

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.

#169235 In pytorch/pytorch;

· lingebeng opened on Nov 29, 2025

[Graph Partition] Wrong graph signature

module: cuda graphsAbility to capture and then replay streams of CUDA kernelsAbility to capture and then replay streams of CUDA kernels

module: inductor

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.

#169232 In pytorch/pytorch;

· BoyuanFeng opened on Nov 29, 2025

[flex_attention] CPU support

Status: Open.

#169224 In pytorch/pytorch;

· vadimkantorov opened on Nov 28, 2025

CPU bad-free in <code>torch.nn.utils.rnn.pad_packed_sequence</code> / <code>unpack_sequence</code>

Status: Open.

#169213 In pytorch/pytorch;

· yyds1233 opened on Nov 28, 2025

Segmentation Fault when Converting QInt32Storage to bfloat16

Status: Open.

#169210 In pytorch/pytorch;

· qiqicliff opened on Nov 28, 2025

Segmentation fault in torch.native_batch_norm with empty running_var tensor

Status: Open.

#169208 In pytorch/pytorch;

· qiqicliff opened on Nov 28, 2025

AOTI autotune failed due to device mismatch

Status: Open.

#169197 In pytorch/pytorch;

· larryliu0820 opened on Nov 28, 2025

[Inductor] cpu triton backend fails with combo kernels on argmax/argmin ops

module: inductor

oncall: pt2

Status: Open.

#169195 In pytorch/pytorch;

· karthickai opened on Nov 27, 2025

dim broadcast in masked tensor between mask and data.

Status: Open.

#169191 In pytorch/pytorch;

· YagaoDirac opened on Nov 27, 2025