Skip to content

Pull requests: oneapi-src/oneDNN

New pull request New
62 Open 916 Closed
62 Open 916 Closed
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
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

xe: sdpa: Improve performance of quantization with better alignment and prefetching performance platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2322 opened Dec 27, 2024 by umar456 Loading…
1
common: verbose: fix acc-mode option dump and softmax diff_dst dump (fixes MFDNN-12921)
#2321 opened Dec 27, 2024 by dzarukin Loading…
2
graph: backend: passes: verbose log enhancement component:graph-api Codeowner: @oneapi-src/onednn-graph
#2320 opened Dec 27, 2024 by rongzha1 Loading…
1
brgemm: support arbitrary K on AMX platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#2319 opened Dec 27, 2024 by ankalinin Loading…
1
graph: backend: dnnl: enhance fp8 matmul patterns component:graph-api Codeowner: @oneapi-src/onednn-graph component:tests Codeowner: @oneapi-src/onednn-arch
#2316 opened Dec 26, 2024 by wzt1997 Loading…
@wzt1997
2
benchdnn: graph: check partition number by default component:graph-api Codeowner: @oneapi-src/onednn-graph component:tests Codeowner: @oneapi-src/onednn-arch documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#2314 opened Dec 24, 2024 by TaoLv Loading…
4
graph: backend: dnnl: encode mem address into constant cache key component:graph-api Codeowner: @oneapi-src/onednn-graph component:tests Codeowner: @oneapi-src/onednn-arch
#2312 opened Dec 24, 2024 by xiang1guo Loading…
@xiang1guo
8
xe: conv: apply skip-w optimization for large strides backport platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2311 opened Dec 23, 2024 by echeresh Loading…
1
src: cpu: aarch64: Re-enable matmul and convolution static quantisation through ACL. platform:cpu-aarch64 Codeowner: @oneapi-src/onednn-cpu-aarch64
#2308 opened Dec 22, 2024 by renato-arantes Loading…
2 tasks done
3
[GPU] Fix jit GEMM debug strategy submission bug A confirmed library bug platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2307 opened Dec 20, 2024 by kealan-barbieri Loading…
1 of 2 tasks
xe: conv: apply skip-w optimization for large strides platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2306 opened Dec 20, 2024 by echeresh Loading…
2
doc: graph: add document for sdpa with compressed key and value documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#2301 opened Dec 20, 2024 by wzt1997 Loading…
@wzt1997
11
src: handle DNNL_VERBOSE=OFF support through an early exit (fixes MFDNN-12862) component:graph-api Codeowner: @oneapi-src/onednn-graph platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64 platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2300 opened Dec 20, 2024 by dzarukin Loading…
3
xe: jit: enable kernel debug information platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2297 opened Dec 19, 2024 by rjoursler Loading…
5
build: cmake: unsupported compilers cleanup documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#2296 opened Dec 19, 2024 by spalicki Loading…
@spalicki
7
tests: benchdnn: add experimental support for sycl graph component:tests Codeowner: @oneapi-src/onednn-arch documentation A request to change/fix/improve the documentation. Codeowner: @oneapi-src/onednn-doc
#2295 opened Dec 19, 2024 by spalicki Loading…
@spalicki
13
x64: brdgmm conv: enable zps per group (backport) backport platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#2292 opened Dec 19, 2024 by tczeszun • Draft
graph: backend: dnnl: support permute for scale and zps inputs component:graph-api Codeowner: @oneapi-src/onednn-graph component:tests Codeowner: @oneapi-src/onednn-arch
#2291 opened Dec 19, 2024 by wzt1997 Loading…
@wzt1997
12
[GPU] Conv and Gemm fix platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2290 opened Dec 18, 2024 by dyoussif Loading…
2
xe: conv: fix tuning support platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2288 opened Dec 18, 2024 by echeresh Loading…
1
[backport v3.7] SDPA fixes for regular scale tensor and alignment for K tensor backport platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2284 opened Dec 18, 2024 by umar456 Loading…
2
rfcs: graph api: Constant Block Weight Mechanism Graph API design RFC A design document
#2280 opened Dec 17, 2024 by xiang1guo Loading…
@xiang1guo
4
xe, benchdnn, x64: Enable large buffer GEMM and Convolution testing for GPU component:tests Codeowner: @oneapi-src/onednn-arch platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64 platform:gpu-intel Codeowner: @oneapi-src/onednn-gpu-intel
#2276 opened Dec 17, 2024 by rjoursler Loading…
2
x64: brdgmm conv: enable zps per group (closes MFDNN-12556) component:tests Codeowner: @oneapi-src/onednn-arch platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#2274 opened Dec 16, 2024 by tczeszun Loading…
10
cpu: x64: brgemm, matmul: add f32:f16 configuration support on AVX2 and AVX512_CORE (fixes MFDNN-11992) component:tests Codeowner: @oneapi-src/onednn-arch platform:cpu-x64 Intel64/AMD64 processors. Codeowner: @oneapi-src/onednn-cpu-x64
#2272 opened Dec 14, 2024 by dzarukin Loading…
8
ProTip! Type g i on any issue or pull request to go back to the issue listing page.