List view

[distributed] AttributeError: module 'torch.xpu' has no attribute '_sleep'

module: distributedFor distributed feature issueFor distributed feature issue

Status: Open.

Feature

#1727 In intel/torch-xpu-ops;

· PenghuiCheng opened on Jun 9, 2025

[distributed] ValueError: Cannot use ReduceOp.PREMUL_SUM with XCCL

module: distributedFor distributed feature issueFor distributed feature issue

Status: Open.

Feature

#1571 In intel/torch-xpu-ops;

· daisyden opened on Apr 11, 2025

[Evaluated] FP8 support in matmul // Issues in test_matmul_cuda.py due to FP8

triaged

Status: Open.

Feature

#322 In intel/torch-xpu-ops;

· yuchengliu1 opened on May 25, 2024

Investigate whether pad mm is useful on XPU

enhancementNew feature or requestNew feature or request

performance

Status: Open.

Task

#1129 In intel/torch-xpu-ops;

· jianyizh opened on Nov 29, 2024

[ARC-WSL-Ubuntu24.04] WSL crash when running torchbench

bugSomething isn't workingSomething isn't working

client

os: WindowsWindows PlatformWindows Platform

Status: Open.

#1502 In intel/torch-xpu-ops;

· libohao1201 opened on Mar 25, 2025

[ARC-WSL-Ubuntu24.04] 15 Timm models got fail_accuracy

bugSomething isn't workingSomething isn't working

client

E2E

module: inductor

os: WindowsWindows PlatformWindows Platform

Status: Open.

#1505 In intel/torch-xpu-ops;

· libohao1201 opened on Mar 25, 2025

Moco NotImplementedError: xpu not supported

Accuracy

dtype: amp_bf16

dtype: amp_fp16

dtype: bfloat16

dtype: float16

dtype: float32

E2E

module: torchbench

training

Status: Open.

#489 In intel/torch-xpu-ops;

· mengfei25 opened on Jun 27, 2024

Timm_efficientdet NotImplementedError: The original model code forces the use of CUDA.

Accuracy

dtype: amp_bf16

dtype: amp_fp16

dtype: bfloat16

dtype: float16

dtype: float32

E2E

module: torchbench

training

triaged

Status: Open.

#492 In intel/torch-xpu-ops;

· mengfei25 opened on Jun 27, 2024

Reduction: Enhance reduction kernel with supporting data type dynamic cast

performance

Status: Open.

Feature

#685 In intel/torch-xpu-ops;

· fengyuan14 opened on Aug 4, 2024

implement torch.linalg.svd xpu backend

module: op impl

Status: Open.

Feature

#1901 In intel/torch-xpu-ops;

· yao-matrix opened on Aug 1, 2025

implement fused rms norm

kernel_optimization

performance

Status: Open.

Feature

#1905 In intel/torch-xpu-ops;

· jianyizh opened on Aug 5, 2025

[cpp extension] Provide a clear error message when using inconsistent oneapi versions.

dependency component: oneAPI

module: build

Status: Open.

Bug

#1649 In intel/torch-xpu-ops;

· ZhaoqiongZ opened on May 9, 2025

[TorchAO] Int4XPULayout Lacks MoE Quant Support

dtype: int4

module: ao

module: quant

Status: Open.

Feature

#1913 In intel/torch-xpu-ops;

· Stonepia opened on Aug 7, 2025

[release/2.8][pt2e] Pytorch 2.8 rc3 int8 accuracy is much lower than 2.7.1

Accuracy

dtype: int8

hw: PVC

Status: Open.

#1801 In intel/torch-xpu-ops;

· mengfei25 opened on Jul 2, 2025

[TorchAO] For the AWQ path, there’s a noticeable gap in LLaMA and Qwen accuracy between BMG and the 4070 Ti.

module: ao

Status: Open.

#1933 In intel/torch-xpu-ops;

· MingxuZh opened on Aug 13, 2025

Precision issues depend on Compiler

dependency component: oneAPI

triaged

Status: Open.

Bug

#1124 In intel/torch-xpu-ops;

· daisyden opened on Nov 28, 2024

torch.xpu.empty_cache() introduces memory leak

module: core

Status: Open.

Bug

#1745 In intel/torch-xpu-ops;

· songhappy opened on Jun 16, 2025

Ask an API to query GPU type(iGPU/dGPU).

dependency component: oneAPI

Status: Open.

Feature

#1722 In intel/torch-xpu-ops;

· xuhancn opened on Jun 5, 2025

implement torch.linalg.cholesky xpu backend

bug_fix_stage2

module: op impl

Status: Open.

Feature

#1936 In intel/torch-xpu-ops;

· jiqing-feng opened on Aug 15, 2025

implement torch.linalg.pinv xpu backend

bug_fix_stage2

module: op impl

Status: Open.

Feature

#1902 In intel/torch-xpu-ops;

· yao-matrix opened on Aug 1, 2025

Index Put and Index Copy deterministic cases failed

bugSomething isn't workingSomething isn't working

Status: Open.

Bug

#1959 In intel/torch-xpu-ops;

· CuiYifeng opened on Aug 25, 2025

segfault with test_fake_crossref_backward_no_amp_to_sparse_xpu_float32

dependency component: driver

module: ut

skippedUsed for temp UT failure to parallel fixUsed for temp UT failure to parallel fix

Status: Open.

#1962 In intel/torch-xpu-ops;

· daisyden opened on Aug 26, 2025

torch._dynamo.exc.InternalTorchDynamoError: TypeError: cannot create weak reference to 'torch.Event' object

module: ut

Status: Open.

Feature

#1969 In intel/torch-xpu-ops;

· shangerxin opened on Aug 27, 2025

torch._dynamo.exc.BackendCompilerFailed: backend='inductor' raised: RuntimeError: CUDA not available

module: ut

Status: Open.

#1970 In intel/torch-xpu-ops;

· shangerxin opened on Aug 27, 2025

assert len(flat_diff_results) > 0 in TestFakeTensorXPU

module: ut

Status: Open.

#1964 In intel/torch-xpu-ops;

· daisyden opened on Aug 26, 2025