Search results

torch.fmax() between MPS tensor and CPU scalar crashes

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

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

module: regressionIt used to work, and now it doesn'tIt used to work, and now it doesn't

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: Closed (completed).

#143933 In pytorch/pytorch;

· by malfet was closed on Dec 28, 2024

· 

2.6.0

RuntimeError when running backward on MPS: "view size is not compatible" with self-attention block

high priority

module: convolutionProblems related to convolutions (THNN, THCUNN, CuDNN)Problems related to convolutions (THNN, THCUNN, CuDNN)

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

module: regressionIt used to work, and now it doesn'tIt used to work, and now it doesn't

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: Closed (completed).

#142344 In pytorch/pytorch;

· by NripeshN was closed on Dec 13, 2024

· 

2.6.0

[MPS] bucketize crashes with scalar input

module: crashProblem manifests as a hard crash, as opposed to a RuntimeErrorProblem manifests as a hard crash, as opposed to a RuntimeError

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: Closed (completed).

#142203 In pytorch/pytorch;

· by skotapati was closed on Jan 2, 2025

· 

2.6.0

Output size of the matrix multiplication is larger than currently supported by the MPS backend: 72250,72250, needs to be less than 2**32 elements

high priority

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: Closed (completed).

#141909 In pytorch/pytorch;

· by gazonk was closed on Dec 17, 2024

· 

2.6.0

MPS Regression when rendering LTXVideo (after pytorch2.4.1)

high priority

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

module: regressionIt used to work, and now it doesn'tIt used to work, and now it doesn't

triage review

Status: Closed (completed).

#141471 In pytorch/pytorch;

· by cchance27 was closed on Dec 1, 2024

· 

2.6.0