Merge 'main' into 'win-ort-main' @ 39e585ff2b #24353
mschofie
merged
515 commits into
win-ort-main
from
mschofie/merge-1.22-current
Make Nuget package pipeline 1ES compliant (#23803)
839d9dcd
Conveting npm packaging pipeline to 1ES (#23767)
9a2e0090
[webgpu] support resize operator (#23780)
cc3f4120
Upgrade React Native to 0.73 (#23575)
40c329ef
Make Nuget CUDA package pipeline 1ES compliant (#23804)
d5742708
[ARM CPU] Fix flaky hgemmb ut (#23814)
7a3810d3
[TensorRT EP] update oss parser to latest (#23710)
000f2c9f
[webgpu] Fix alignment issues in shader code (#23776)
c6664e20
upgrade emsdk to 4.0.4 (#23819)
6df0973e
[OVEP] Update support for Contrib Ops (#23789)
17f39475
Update onnxruntime_external_deps.cmake: add missing EXCLUDE_FROM_ALL …
b1f2a3f5
Quant tool: Add `nodes_to_exclude` in `get_qnn_qdq_config` (#23779)
5ab953cb
[ORT/CI_Pipeline] Use --enable_generic_interface in ORT builds for EP…
05642657
Increase npm package pipeline ReactNative_CI_iOS timeout to 120 mins …
a189bfca
[Mlas] Unblock hardcoded matmul blocking size (#23815)
c61a4b11
Revert changes onn mac-react-native-ci-pipeline.yml (#23845)
2a4cfab4
Fix flash attention for GQA (Phi4) (#23850)
1be64f88
Model Builder API (#23223)
1088a1ed
Fix typo: change `Upample` to `Upsample`. (#23838)
1ffe793a
[doc] Fix typos in csharp/src/Microsoft.ML.OnnxRuntime/ (#23848)
0a6b05fb
Quant tool: Consistent `get_qdq_config` and `get_qnn_qdq_config` beha…
daf9565d
Change the logic to generate the default ep context file name (#23788)
99c51a32
Make Nuget QNN package pipeline 1ES compliant (#23805)
7f0c2c64
[js/common] allows using Uint16Array as data for float16 tensor (#23827)
18725277
[js/webgpu] Reland the optimization of ConvTranspose (#23858)
325ee309
[OpenVINO] Fix a build warning (#23877)
30c68254
Change gsl::byte to std::byte (#23872)
bde4fbec
Allow using extended minimal build for several EPs (#23834)
17dcea7a
Add dawn to ThirdPartyNotices (#23876)
813bdaab
Enable QNN EP weight sharing generation using public API (#23702)
9d0dc9f0
[QNN-EP]: Fix inference failures while running with htp_shared_memory…
788ca51b
Fix enable_pix_capture build for WebGPU (#23857)
8aed9208
[WebGPU-EP Native] Add ReduceMean (#23860)
834adde8
[WebGPU EP] introduce BiasAdd contrib op (#23861)
cfb0a72f
Dynamo export and improve benchmark script for SAM2 encoder (#23887)
5e636a67
[js/web] improve workaround for bundlers (#23902)
aafa8d17
[webgpu] Restore MatMulNBits workgroup size for Phi-3.5 (#23349)
d35db9b8
[webgpu] support Pad operator (#23141)
95225dda
[WebNN] Accept Float16Array for float16 data type if it is available …
b5242293
Ensure that the 'cmake_minimum_required' is version 3.5 or greater (#…
996fffbe
WebGPU: Remove deprecated subgroups-f16 from WebGPU native and JS EP …
54b2d64c
[JSEP/WebGPU] Fixed error in softmax dispatch. (#23906)
ccf8fdd9
enable WebGPU EP in WebAssembly build (#23913)
101353cf
Adding OpenVINO Windows CI Pipeline (#23919)
8f077435
[WebGPU EP] SoftMax Implementation (#23538)
4bb79d13
Exclude MAUI projects from GPU C# packaging builds (#23923)
b2ab87e8
Support all block sizes that are multiples of 32 for DP4A (#23907)
eeaf73b3
Example custom op with output type inferencing (#23916)
c28bf788
Enabling L2+ Optimizations for EPs (#23517)
1199dc08
fix binplace file in web pipeline (#23930)
2ba076aa
Updated run_CIs_for_external_pr.py to support the Windows OpenVINO CI…
e47c6c16
Fix ConvInteger handling of optional inputs. (#23935)
8969ee78
Updated ov version in pipeline (#595) (#23882)
26f590b3
[AIX] External data handling (#23859)
f25deaea
Create a packaging pipeline for a custom nuget package (#23918)
593d5c0e
Fix license in example test code. (#23936)
7dbbfe08
replace usage of gsl::narrow and gsl::narrow_cast in WebGPU EP (#23926)
ab38607d
VCPKG improvement: set VCPKG_OSX_DEPLOYMENT_TARGET (#23933)
cffef2e0
Allow using a different version of flatbuffers when building with vcp…
49328fe6
Make python package pipeline 1ES compliant (#23800)
95dcd150
Delete ROCM Nuget Publishing Pipeline (#23948)
989d4177
![dependabot[bot]](https://avatars.githubusercontent.com/in/29110?v=4&s=60)
Bump SixLabors.ImageSharp from 2.1.9 to 2.1.10 in /csharp/sample/Micr…
fe7634eb
Make python CUDA package pipeline 1ES compliant (#23802)
246c2191
Migrate yarn to npm (#22116)
773bb4ff
[WebGPU/JSEP] Support group query attention do_rotary attribute (#23524)
333fbdb4
Fix npm audit in js/react-native/e2e (#23975)
f18e9faa
Suppress some warnings in WebGPU EP generated by GCC 13 (#23984)
64436265
Fix NPM audit in js/react-native (#23974)
d010acb5
Bump axios from 1.7.9 to 1.8.2 in /js/node (#23963)
9118b1de
GCC 14: fix insert_or_assign() call (#23955)
5672cf7d
ADD emsdk env vars to VCPKG_KEEP_ENV_VARS (#23997)
d2bf9a79
Fix ONNX Runtime Python Test Pipeline (#23990)
fe435371
[webgpu] Fix the continuation issue (#23999)
16d6f397
[WebGPU EP] Implements Gelu, BiasSplitGelu, and QuickGelu (#23981)
9891eb3d
[Native WebGPU] Added ReduceMax and ReduceSum (#23934)
6dd6ef93
Convert Windows CPU CI Pipeline to Github Actions (#23996)
47bd0468
[Fix] Dependencies find_package Eigen error (#23939)
06482c26
Update onnxruntime_c_api.h to work with MinGW (#24006)
5e057292
Add DNNL github workflow (#24011)
57ddd026
Qnn weight sharing improvement (#23945)
7ae606f7
Correct generated cmake syntax (#24016)
11216a4e
[webgpu] allow to specify UseIndicesTypeAlias for Indices (#24019)
1362e7ca
[webgpu] allow overloads to Program::AddIndices (#24021)
401f24a8
fix test for RotaryEmbedding (#24022)
219c919c
Fix attention bias broadcast (#24017)
99b78a94
Remove unused parameter in csharp InferenceTest (#24031)
5bd31636
[TensorRT EP] Call cudaSetDevice at compute function for handling mul…
6bb6d791
Increase timeout for ARM64-Xcode16-targeting-iphonesimulator (#24030)
3f71d637
Support tvOS build (#24000)
1fc6d8ca
[TensorRT EP] Stop enforcing oss parser during Windows debug build (#…
cb3f631f
Set CMAKE_POLICY_DEFAULT_CMP0069 to NEW to ensure that IPO flags are …
9a296a0a
Make Cuda packaging pipeline 1ES compliant (#23806)
9f214561
[webgpu/wasm] allow runtime switch between WebGPUEP and JSEP (#24032)
7c05e7f5
Move call to MLAS_CPUIDINFO::GetCPUIDInfo() out of MlasSQNBitGemmDisp…
c9c8b48d
[webgpu] fix the wrong dispatch size in flash_attention (#24020)
cc5840be
avoid copy unnecessary files for nodejs pkg (#23992)
41c239df
Add support for custom position ids and attention bias to GQA CPU ope…
5a694bcb
[WebNN] Better int64 integration (#23831)
73d9826a
Convert Windows GPU pipelines and Windows OpenVino pipeline to Github…
b8966665
[ARM CPU] Fix fp16 const initialization on no-fp16 platform (#23978)
f22ee08f
[Native WebGPU EP] Add packedQKV and do_rotary attribute support to G…
ae501eeb
Whisper Redesigned Solution (#23549)
7942fa7a
Windows: Show more useful DLL load errors to say exactly what DLL is …
5ef0d211
Extend CMAKE_CUDA_FLAGS with all Blackwell compute capacity (#23928)
2bc73ca8
[WebGPU] Reduce staging buffers for uploading intializers (#23968)
f5812d0e
[WebGPU EP] Implement Remaining Reduction Ops (#24045)
154e3b7d
add bool support to EPContext schema to unblock some models (#24065)
a46d2127
[WebGPU EP] fix for reduce min/max error on MacOS CI (#24077)
b3aa5a3c
Upgrade current MacOS-13 to 14 (#23293)
e495750a
Fix CUDA EP Abs and Sign bfloat16 support (#23914)
c6a26754
Improve typing for OrtValue and other public Python interfaces (#24086)
12fea572
[webgpu] Limit that K must be divisible by 128 to apply dp4a matmul (…
a85977dd
Add macOS ARM64 pipeline for webgpu (#24060)
d98046b3
[WebNN/WebGPU JS] Fix shared Module methods overriding each other (#2…
eceae8b2
Enable multithreading on FP16 to FP32 cast operator (#23619)
7fc7d5ec
Move Android CI Pipeline to Github Actions (#24094)
3488ba39
Cleanup CoreML EP's code to remove COREML_ENABLE_MLPROGRAM (#23490)
7444feeb
webgpu ep support for argmax/argmin (#24089)
b626409e
[mobile/reactnative] Remove namespace from AndroidManifest.XML to res…
d8ed4da1
[WebGPU EP] fix implementation of Pow (#24088)
80441e4e
Increase timeout to 90min for ARM64-Xcode16-targeting-iphonesimulator…
731b27e2
[WebGPU] fix test failure in Reduce operators on macOS ARM64 (#24108)
da7874c8
[WebGPU EP] Implements CumSum Operator (#24047)
8d21bf72
[webgpu] Use 1d dispatch group size (#24084)
81a89204
[WebGPU] fix test failure in MatMulNBits on macOS ARM64 (#24109)
9dcb99cd
[QNN-EP] Add support for Sum operator with 2 inputs (#24098)
4d5e274f
[WebNN] Replace narrow with SafeInt for consistently in integer handl…
5d43f0ab
[QNN-EP] Add Lora Support with offline QNN context binary (#24026)
6bdbf08c
[TensorRT EP] support TensorRT 10.9-GA (#23905)
440d17a7
[webgpu] Apply dp4a for generation shader (#24064)
127c8503
[CUDA] Support slide window in cutlass fused attention (#24072)
db0c95c1
[MIGraphX EP] rename HIPPinnedAllocator to MIGraphXPinnedAllocator (#…
16b0b323
[MIGraphX EP] check POLICY CMP0144 availability before used (#24104)
9922d480
[JSEP] handles edge case in gridsample operator (#24121)
469fb7e3
[OpenVINO]Session Options Appended After AppendExecutionProvider (#23…
49024a1e
[webgpu]Add MaxPool and AveragePool (#23714)
7a6514c8
[webgpu EP] put GetMaxComponents and SumVector to one place. (#24122)
9e53afab
skip MOE python test when MPI is not installed (#24116)
dcc1f5ac
Integrate KleidiAI for MatMulNBits via MlasQNBitGemm (#23627)
90c5ffb5
add test cases for webgpu ep in web (#24117)
0a363d9e
Refactor Webnn IsSupported*() to use constant initializers. (#24118)
cd9406bf
Deleted the constant SKIP_CUDA_TEST_WITH_DML (#24113)
4959468a
Update T5 Onnx Export and Optimization (#23949)
d84314cb
Update package.json to make the dist avaliable again (#23991)
3012d445
Fix attention QK linkage error (#24134)
2b3d7fb1
Bump next from 15.1.2 to 15.2.3 in /js/web/test/e2e/exports/testcases…
5ed900e9
[Shape Inference] Add shape inference for QLinearAdd and QLinearMul o…
2b5c9da6
[mobile] Add Android NuGet BrowserStack test to NuGet packaging pipel…
8eb8c2b0
[CPU] Add fp16 support to sparse attention (#24015)
828e3726
refactor mac CI pipelines (#24138)
373b9e2a
Address Windows CUDA build issue (#24149)
5244d68b
[webgpu] add option to perserve device and enable in unittest (#24115)
e03631ee
[js/web] allow bundler import condition for not bundling wasm (#24014)
78d91cdd
[js] Add API for accessing metadata of a model's input/output (#23937)
618aef7e
add cache "onnxnodetests" for node tests (#24150)
afaf4a5e
[Native WebGPU] Add Matmul (#24046)
ce65e253
Upgrade Big Model pipeline CUDA from 11.8 to 12.x (#24156)
bb005b93
Proper Error Message when fp16 model is used for Beam Search in CPU (…
de502c89
Change type len from int to size_t (#24157)
a4b8f11c
Limit the Pipeline ability to build cuda 11 (#24073)
a8fb7868
Move Linux CPU CI pipeline to Github Actions (#24154)
86806677
Bump vite from 6.2.1 to 6.2.3 in /js/web/test/e2e/exports/testcases/v…
d9c961ce
[onnxruntime_perf_test] Fix custom_allocator_ destruction order. (#24…
1ef30446
Fix layout transformer for FusedConv (#24169)
25b06f20
Migrate Zip-Nuget Package Pipeline to 1ES (#23609) Also, kleidail is …
1f6dc881
Update the min GCC version (#24148)
9dbfee91
[QNN EP] ARM64EC python package remove --vcpkg in build (#24174)
2a800d1e
[WebGPU EP] Add GEMM implementation (#24023)
a8673c6e
[wasm] remove --vcpkg in wasm build (#24179)
513e8de1
revise mac os pipeline to reduce the amount of jobs (#24177)
32b376cd
fix triggering for "Validate Gradle Wrapper" pipeline (#24181)
be1cfc4e
upgrade QNN to version 2.32.0.250228 (#23977)
5d805c23
[JSEP] adjust edge case logic for scatternd (#24172)
24ece479
Make the custom nuget packaging pipeline 1ES commpliant. (#24191)
1f70fc25
Disable KleidiAI in Python Packaging pipeline MacOS build (#24194)
4d13b70f
Rolling back the python/cuda (#24170)
041674ad
Remove all CG template from pipelines (#24193)
914be22e
Move Linux ARM64 CI pipeline and Linux DNNL CI pipeline to Github Act…
bd00c39f
[webgpu-ep] Fix test_batchnorm_example (#24184)
86b4c789
Further reduce work load for Mac CI pipeline (#24197)
26566710
Generate unique names for SliceSplit fusion. (#24217)
64b0d071
Fix the pipeline that failed because of vcpkg (#24226)
25921476
Improve Shape Inference for GQA (#24143)
c756e0ab
Add React Native namespace back in for iOS (#24218)
19d8d69c
RoPE fp16 avx (#23772)
180ba8f8
Migrate Linux GPU pipelines to Github Actions (#24232)
f430dce9
Migrate Web CI into github actions (#24219)
41dde351
update the readme doc for the tool ep_weight_sharing_ctx_gen (#24233)
4a669fd1
[WebGPU EP] If Implementation for WebGPU EP (#24242)
7ef0ddc5
Update linux-dnnl.yml: rename the pipeline (#24240)
8de342ad
[webgpu] Fix test_layer_normalization_2d_axis0 (#24223)
d71aa4d8
[webgpu] fix LayerNorm with empty input (#24244)
f1d790c2
Bump actions/setup-python from 4 to 5 (#24251)
492af7a3
Bump actions/cache from 3 to 4 (#24250)
83650edc
[QNN EP] Add platform-agnostic EP option to specify QNN backend, `bac…
22787aec
[webgpu] Fix opset-12 softmax nhwc issue (#24227)
ad2e5652
Extend pyright exclude list in pyproject.toml (#24246)
528f29a8
[js/web] Add Wasm Relaxed SIMD support to wasm backend (#22794)
ba2999c5
Add shader key validation step in WebGPU CI pipeline (#24243)
4eeefd72
upgrade dawn version to 4cb1f9be152a4fa6bb695c08cd707ab078a1e2fb (#24…
30115cfe
Bump dsaltares/fetch-gh-release-asset from 1.1.0 to 1.1.2 (#24248)
5982430a
Bump vite from 6.2.3 to 6.2.4 in /js/web/test/e2e/exports/testcases/v…
e2274150
[WebGPU EP] fixes bugs in split implementation (#24259)
5068ab9b
Bump microsoft/onnxruntime-github-actions from 35f8bd42417991aa46577e…
1b48cc41
Update xcode and iphoneSimulatorVersion after MacOS-14 (#24260)
5b080558
Exclude onnxruntime-inference-examples directory from Component Gover…
24620e70
[VitisAI] Fixed include error. (#24199)
67216c89
Migrate pull:wasm to github action (#24269)
a5bc69c5
Ensure to use correct GPU device in RunSince when it's invoked by new…
b3793906
Adding build-system to pyproject.toml (#24216)
b5d15bc9
[WebGPU EP] Implements ceil mode for Average Pool (#24270)
bc7b07db
Pin vcpkg version (#24284)
55aa03c1
Support load TensorRT V3 plugin (#24211)
a14d586d
Expose TRT preview features as EP option (#24212)
21db38c1
[webgpu] test_layer_normalization_3d_axis0_epsilon (#24276)
8465ca38
[webgpu][dawn API optimization] reduce number of calls to wgpuDeviceH…
7a551887
Bump next from 15.2.3 to 15.2.4 in /js/web/test/e2e/exports/testcases…
d2388135
Bump image-size from 1.1.1 to 1.2.1 in /js/react_native/e2e (#24278)
cbaa8bc7
[QNN-EP] Enhance QNN-EP support for Softmax with opset < 13. (#24180)
a28da4b9
Update publish-nuget.yml to correct feed. (#24299)
e5e906ee
[webgpu] Optimize MatMulNBits for f16 Block32 prefill performance (#2…
3dfc2ae3
upgrade action shellcheck to v1.30.0 (#24304)
82c8e569
[QNN-EP] Fix ONNX context model helper. (#24271)
1cb53d00
[WebGPU] fix Pad cache key (#24305)
318cc87f
Bump vite from 6.2.4 to 6.2.5 in /js/web/test/e2e/exports/testcases/v…
56f10183
[WebGPU] fix cache key of AttentionProbs/VxAttentionScore (#24309)
2e94c5a4
Support Gemma3 with Clip fused attention (#24280)
e944379e
Update packaging pipeline for Nodejs binding (#24301)
11fda2ad
Add support for uint8_t as data type for GatherBlockQuantized (#24239)
a4976e33
[Native WebGPU] Add Conv, ConTranspose and FusedConv (#24186)
9102aaee
[webgpu][dawn API optimization] reduce number of calls to wgpuDeviceG…
a7e62d63
Fix 'minimal_power' to 'minimum_power' for DirectML performance selec…
55c1a3b0
Add ConvTranspose cache key (#24317)
d6df4f29
[webgpu] Use 1D dispatch groups for attention (#24228)
a1186f63
[webgpu][dawn API optimization] reduce number of calls to buffer APIs…
73676fc5
Implement load cancellation ability (#24257)
350d1400
[webgpu] Fix ROUND_PREFER_CEIL issue of Resize operator (#24229)
ca1b32df
[Native WebGPU] Exclude WebGPU EP from ConvFp16 3D tests. (#24327)
b803429a
[VitisAI EP] export InferShapes to VitisAIEP (#23881)
554fb4ad
[webgpu] Flash attention for generation (#23808)
18f91e55
Use WASM f32x4 relaxed min/max for relaxed simd build (#24324)
04e0b50c
webgpu support for DequantizeLinear (#24268)
f83e6618
[webgpu] fix the reflect mode issue of Pad (#24202)
10e51d26
Remove explicit batch network flag for TRT 10+ (#24298)
4edada60
[webgpu] Fix bias_split_gelu (#24342)
22656134
[webgpu] fix bias-add (#24336)
34abb8b4
[webgpu] optimize SkipLayerNormalization operator (#24164)
0acb0488
ROCm: Remove -Wno-interference-size compiler flag (#24326)
d7a38a57
[web] revise flag `ort.env.wasm.simd` (#24314)
39e585ff
Merge 'main' into 'win-ort-main' @ "39e585ff2b:[web] revise flag `ort…
529a34e6
Login to write a write a comment.
Login via GitHub