Search results

Use opmath_t and not double compute in fused optimizers

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

module: optimizerRelated to torch.optimRelated to torch.optim

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

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.

#153649 In pytorch/pytorch;

· janeyx99 opened on May 15, 2025

Support int step for nonfused optimizer implementations

actionable

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: optimizerRelated to torch.optimRelated to torch.optim

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.

#142378 In pytorch/pytorch;

· janeyx99 opened on Dec 9, 2024

Adafactor foreach impl performance tracker

actionable

module: mtaIssues related to multi-tensor apply kernels and foreach functionsIssues related to multi-tensor apply kernels and foreach functions

module: optimizerRelated to torch.optimRelated to torch.optim

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

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.

#133367 In pytorch/pytorch;

· janeyx99 opened on Aug 13, 2024

Adafactor compile support tracker

module: optimizerRelated to torch.optimRelated to torch.optim

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.

#133268 In pytorch/pytorch;

· janeyx99 opened on Aug 12, 2024

Optimizer step pre/post hook tests should try harder to ascertain the hook happens before/after the step call

actionable

better-engineeringRelatively self-contained tasks for better engineering contributorsRelatively self-contained tasks for better engineering contributors

module: optimizerRelated to torch.optimRelated to torch.optim

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.

#119300 In pytorch/pytorch;

· janeyx99 opened on Feb 6, 2024

Add support for sparse tensors with view_as_real

actionable

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

module: optimizerRelated to torch.optimRelated to torch.optim

module: sparseRelated to torch.sparseRelated to torch.sparse

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.

#118153 In pytorch/pytorch;

· janeyx99 opened on Jan 24, 2024

Enable all optimizers (except SparseAdam + LBFGS) to be capturable

actionable

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

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

module: optimizerRelated to torch.optimRelated to torch.optim

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.

#118018 In pytorch/pytorch;

· janeyx99 opened on Jan 22, 2024

LBFGS MPS test_can_load_older_state_dict fails on older MacOS versions

module: macosMac OS related issuesMac OS related issues

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

module: optimizerRelated to torch.optimRelated to torch.optim

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

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.

#117094 In pytorch/pytorch;

· janeyx99 opened on Jan 10, 2024

Support float64 step in fused optimizers

actionable

module: optimizerRelated to torch.optimRelated to torch.optim

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.

#116592 In pytorch/pytorch;

· janeyx99 opened on Jan 1, 2024

Add BC test for load_state_dict on OptimizerInfos and ModuleInfos

better-engineeringRelatively self-contained tasks for better engineering contributorsRelatively self-contained tasks for better engineering contributors

module: nnRelated to torch.nnRelated to torch.nn

module: optimizerRelated to torch.optimRelated to torch.optim

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

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.

#116502 In pytorch/pytorch;

· janeyx99 opened on Dec 28, 2023

Legitimize/harden differentiable optimizers in torch.optim

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

module: optimizerRelated to torch.optimRelated to torch.optim

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.

#116490 In pytorch/pytorch;

· janeyx99 opened on Dec 28, 2023

Refactor precision handling in test/test_optim.py

actionable

better-engineeringRelatively self-contained tasks for better engineering contributorsRelatively self-contained tasks for better engineering contributors

module: optimizerRelated to torch.optimRelated to torch.optim

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.

#116203 In pytorch/pytorch;

· janeyx99 opened on Dec 20, 2023