Search results

RFC move to Pyrefly for Type Checking

module: typingRelated to mypy type annotationsRelated to mypy type annotations

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#163283 In pytorch/pytorch;

· maggiemoss opened on Sep 18, 2025

autograd assertion error in cuda backend when privateuse1 is renamed

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

module: PrivateUse1private useprivate use

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#161129 In pytorch/pytorch;

· matthewdcong opened on Aug 21, 2025

Offer official Pytorch Vulkan backend on pytorch.org

featureA request for a proper, new feature.A request for a proper, new feature.

module: vulkan

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#160230 In pytorch/pytorch;

· alshdavid opened on Aug 8, 2025

Add nn.GradBank for gradient scaling to prevent vanishing/exploding gradients

module: nnRelated to torch.nnRelated to torch.nn

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#159765 In pytorch/pytorch;

· danindiana opened on Aug 4, 2025

Enable DataLoaders to prefetch batches across epochs

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

module: dataloaderRelated to torch.utils.data.DataLoader and SamplerRelated to torch.utils.data.DataLoader and Sampler

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#158704 In pytorch/pytorch;

· Giuspepe opened on Jul 19, 2025

Compatibility with C++20 <code>std::format</code>

module: buildBuild system issuesBuild system issues

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#158346 In pytorch/pytorch;

· BwL1289 opened on Jul 15, 2025

Add GoLU Activation Function 🚀

module: nnRelated to torch.nnRelated to torch.nn

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#158043 In pytorch/pytorch;

· IndrashisDas opened on Jul 10, 2025

research adding cuda-bindings to core

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

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#156424 In pytorch/pytorch;

· msaroufim opened on Jun 19, 2025

Vulkan interoperability

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

module: vulkan

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#155986 In pytorch/pytorch;

· cajoek opened on Jun 14, 2025

as_tensor of list of tensors should keep grad history

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

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#155983 In pytorch/pytorch;

· danra opened on Jun 14, 2025

RFC: Add Adaptive Entropy-Gated Contrastive Fusion (AECF) for Robust Multimodal Attention Pooling

module: nnRelated to torch.nnRelated to torch.nn

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

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#155878 In pytorch/pytorch;

· leochlon opened on Jun 13, 2025

Change error to warnning when tensor size mismatch while loading params

has workaround

module: nnRelated to torch.nnRelated to torch.nn

needs researchWe need to decide whether or not this merits inclusion, based on research worldWe need to decide whether or not this merits inclusion, based on research world

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.

#155368 In pytorch/pytorch;

· PistonY opened on Jun 6, 2025