Skip to content

Pull requests: intel/torch-xpu-ops

Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Assignee
Filter by who’s assigned
Assigned to nobody Loading
Sort

Pull requests list

[WIP] Initialize total_weight tensor with zeroes in nll_loss2d_forward_kernel disable_accelerate Disable accelerate test job in PR CI testing disable_distributed Disable distributed UT test jobs for the PR disable_e2e Disable all e2e test jobs for the PR disable_transformers Disable transformers UT test in PR CI disable_win Disable Windows CI test jobs for the PR
#3294 opened Apr 9, 2026 by BBBela Contributor Draft
[sparse][xpu] Add efficient CSR sparse addition kernel
#3292 opened Apr 9, 2026 by tszulist-hbn Contributor Loading…
Fix asarray requires_grad expectations in XPU tests
#3291 opened Apr 9, 2026 by BartoszKokoszko Contributor Loading…
Fix FlashAttention meta registeration issue
#3289 opened Apr 9, 2026 by LuFinch Contributor Loading…
skip won't fix cases in #2309, #2164, #2536, #2472
#3288 opened Apr 9, 2026 by daisyden Contributor Loading…
Enable dpclang sycl compiler
#3287 opened Apr 8, 2026 by dvrogozh Contributor Loading…
Remove stale ConvTranspose3d complex half parity xfail
#3285 opened Apr 8, 2026 by gplutop7 Contributor Loading…
Override test case test_big_num_tensors to change torch.cuda to torch.xpu. disable_accelerate Disable accelerate test job in PR CI testing disable_distributed Disable distributed UT test jobs for the PR disable_e2e Disable all e2e test jobs for the PR disable_transformers Disable transformers UT test in PR CI disable_win Disable Windows CI test jobs for the PR
#3283 opened Apr 8, 2026 by BBBela Contributor Loading…
Add contiguous gather fast path bypassing OffsetCalculator
#3282 opened Apr 8, 2026 by yucai-intel Contributor Loading…
[DO NOT MERGE][TEST ONLY] Create AGENTS.md disable_all Disable all ci test jobs for the PR, just keep basic lint check
#3281 opened Apr 8, 2026 by Stonepia Contributor Draft
Optimize scatter kernel launch with compile-time constants
#3279 opened Apr 8, 2026 by yucai-intel Contributor Loading…
Use C++20 for SYCL device compilation
#3278 opened Apr 7, 2026 by r-barnes Contributor Loading…
Use std::fma in addcmul and foreach pointwise ops for FMA parity with CUDA
#3275 opened Apr 7, 2026 by AKloniecki Contributor Loading…
3 tasks done
[CI] Seperate Windows CI disable_distributed Disable distributed UT test jobs for the PR disable_e2e Disable all e2e test jobs for the PR disable_ut Disable all non-distributed UT test jobs for the PR
#3274 opened Apr 7, 2026 by RUIJIEZHONG66166 Contributor Loading…
Fix bf16/fp16 accuracy issue in sparsecsr addmm
#3273 opened Apr 6, 2026 by jenniew Contributor Loading…
[CI] Fix permission issue for workflows disable_win Disable Windows CI test jobs for the PR
#3271 opened Apr 6, 2026 by chuanqi129 Contributor Loading…
Migrate std::exp to sycl::exp across XPU kernel files
#3269 opened Apr 6, 2026 by jianyizh Contributor Loading…
Decompose dot_xpu_mkl into mul and sum for long dtype
#3265 opened Apr 3, 2026 by Silv3S Contributor Loading…
Add native XPU SVD implementation using oneMKL gesvd
#3264 opened Apr 3, 2026 by PatrykWilczewski Contributor Loading…
Improve norm kernel performance for 16-bit types on XPU
#3263 opened Apr 3, 2026 by yucai-intel Contributor Loading…
Add triangular solve function for sparse CSR tensors in XPU
#3261 opened Apr 2, 2026 by tszulist-hbn Contributor Loading…
Skip TypedStorage tests on XPU
#3260 opened Apr 2, 2026 by BartoszKokoszko Contributor Loading…
ProTip! Exclude everything labeled bug with -label:bug.