Search results

Missing absolute tolerance transform for complex inputs breaks slow/fast gradcheck contract

module: autogradRelated to torch.autograd, and the autograd engine in generalRelated to torch.autograd, and the autograd engine in general

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

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.

#166385 In pytorch/pytorch;

· redwrasse opened on Oct 28, 2025

[inductor] Convert complex operations to real-block matrices

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

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.

#163984 In pytorch/pytorch;

· Marius-Juston opened on Sep 26, 2025

torch.arange not implemented for complex numbers

enhancementNot as big of a feature, but technically not a bug. Should be easy to fixNot as big of a feature, but technically not a bug. Should be easy to fix

has workaround

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

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.

#160629 In pytorch/pytorch;

· springcrane opened on Aug 14, 2025

<code>torch.nn.functional.sigmoid</code> produces inconsistent results on different device for complex inputs

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

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

module: NaNs and InfsProblems related to NaN and Inf handling in floating pointProblems related to NaN and Inf handling in floating point

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.

#159870 In pytorch/pytorch;

· rookieLiu2018 opened on Aug 5, 2025

<code>torch.compile</code> fails on <code>torch.vdot</code> with complex tensors

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

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.

#157607 In pytorch/pytorch;

· raindrops-0199 opened on Jul 4, 2025

<code>torch.reciprocal</code> and <code>torch.divide</code> for Complex <code>inf</code> Incorrectly Returns <code>NaN</code> Only for Tensors with &gt;= 4 Elements on CPU

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

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

module: edge casesAdversarial inputs unlikely to occur in practiceAdversarial inputs unlikely to occur in practice

module: NaNs and InfsProblems related to NaN and Inf handling in floating pointProblems related to NaN and Inf handling in floating point

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.

#157272 In pytorch/pytorch;

· rookieLiu2018 opened on Jun 30, 2025

torchscript: RuntimeError Unsupported dtype complex

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

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

Status: Open.

#156814 In pytorch/pytorch;

· fzimmermann89 opened on Jun 25, 2025

Inconsistent <code>torch.rsqrt</code> results on complex128 between CPU and CUDA

low priorityWe're unlikely to get around to doing this in the near futureWe're unlikely to get around to doing this in the near future

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

module: edge casesAdversarial inputs unlikely to occur in practiceAdversarial inputs unlikely to occur in practice

module: NaNs and InfsProblems related to NaN and Inf handling in floating pointProblems related to NaN and Inf handling in floating point

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.

#156152 In pytorch/pytorch;

· rookieLiu2018 opened on Jun 17, 2025

<code>torch.cumprod</code> for complex128 produces inconsistent results between CPU and CUDA

low priorityWe're unlikely to get around to doing this in the near futureWe're unlikely to get around to doing this in the near future

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

module: edge casesAdversarial inputs unlikely to occur in practiceAdversarial inputs unlikely to occur in practice

module: NaNs and InfsProblems related to NaN and Inf handling in floating pointProblems related to NaN and Inf handling in floating point

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.

#154606 In pytorch/pytorch;

· rookieLiu2018 opened on May 29, 2025

<code>torch.add</code> and <code>torch.sub</code> return incorrect results when computing complex numbers containing <code>inf</code> on both CPU and GPU

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

module: NaNs and InfsProblems related to NaN and Inf handling in floating pointProblems related to NaN and Inf handling in floating point

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.

#154501 In pytorch/pytorch;

· rookieLiu2018 opened on May 28, 2025

torch.fft.irfft(n) doesn’t handle non-Hermitian inputs in a consistent way

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

module: fft

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.

#154496 In pytorch/pytorch;

· pavel-shmakov opened on May 28, 2025

torch.jit.script gives false results for autograd if complex data types are involved

module: complexRelated to complex number support in PyTorchRelated to complex number support in PyTorch

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

Status: Open.

#154324 In pytorch/pytorch;

· frank-roesler opened on May 25, 2025