diff --git "a/Microsoft_onnxruntime.json" "b/Microsoft_onnxruntime.json"
new file mode 100644--- /dev/null
+++ "b/Microsoft_onnxruntime.json"
@@ -0,0 +1 @@
+[{"node": {"oid": "06269a3952fb1759d93235b9d66f9beb10ae8663", "messageHeadline": "[js/webgpu] allow uint8 tensors for webgpu (#19545)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "4874a41008138ecc1f26e9cd17e5d9d7febb29aa", "messageHeadline": "[QNN EP] Update default QNN SDK to 2.19.2.240210 (#19546)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "44d8ad93b20efdba921ca80f23485c084b5174d0", "messageHeadline": "Whisper Timestamps and Temperature (#19509)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 560, "deletions": 379}}, {"node": {"oid": "1dce5e17321d50bf345022b525a937933473415a", "messageHeadline": "Disable TF32 in Linux_Test stage of Linux GPU CI Pipeline (#19541)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b84712151c06f0f59359916be572f71bd36721a4", "messageHeadline": "QNN EP: Fuse DQ -> Q sequences into a QNN Convert op (#19511)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 319, "deletions": 41}}, {"node": {"oid": "ef0b71308c0e2395d3ea63e627515ff8e624ad45", "messageHeadline": "Optimize KahnsTopologicalSort and PriorityNodeCompare (#19475)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 85, "deletions": 45}}, {"node": {"oid": "4bfa69def85476b33ccfaf68cf070f3fb65d39f7", "messageHeadline": "Speed Up DecoderMaskedSelfAttentionTest (#19531)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 242, "deletions": 209}}, {"node": {"oid": "d0061d6fb15d40eeb35fa1b40a414cd231d51db9", "messageHeadline": "Update stale.yml to use old version as a bug fix (#19532)", "author": {"name": "sophies927", "email": "107952697+sophies927@users.noreply.github.com", "user": {"login": "sophies927"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d63c664ca0021fbac31cee57ff1eaa8bce3d1903", "messageHeadline": "fix rocm  ci pipeline (#19525)", "author": {"name": "rui-ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "660f39aca5d47888804163405c64ee67eec6eed5", "messageHeadline": "Perf improvement for Intel MTL CPUs (#19524)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 162, "deletions": 7}}, {"node": {"oid": "775c774f4bdcdd57c107030e1341809b4b5ba35e", "messageHeadline": "Add BF16 to Sqrt (#19363)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 17, "deletions": 4}}, {"node": {"oid": "a67e6925468effd1897c2f541821d32a2860a037", "messageHeadline": "add GatherSliceToSplitFusion and Unittest (#19218)", "author": {"name": "rui-ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 519, "deletions": 0}}, {"node": {"oid": "4e5119760d8cf1c2e751f4264f23ab3e5a25aebc", "messageHeadline": "Add initial support for CoreML ML Program to the CoreML EP. (#19347)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2131, "deletions": 1037}}, {"node": {"oid": "944d8f85135e0caf836ae7f6ad1bfac8dcba2f21", "messageHeadline": "Update the default std flag used during torch extensions compilation \u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3b03b2e046092522e84f0b9aebac1b394a3e4b13", "messageHeadline": "Upgrade default ORTModule opset from 15 to 17 (#19315)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 51, "deletions": 27}}, {"node": {"oid": "1508c2ee39023274417417b290303cf058ceedd6", "messageHeadline": "Restrict L2 Cache Core check to Intel devices (#19483)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 21, "deletions": 4}}, {"node": {"oid": "fbff99a432caef529f90d20137fa5aee33f38fcf", "messageHeadline": "Change Jave Test Threshold (#19508)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f53d2c2465d81cdb4e14c7241eab327184192c88", "messageHeadline": "Phi2 script fixes (#19500)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 62, "deletions": 46}}, {"node": {"oid": "544407038d96521617fe633cf97153d3e75561f5", "messageHeadline": "SimplifiedLayerNormalization Fusion BFloat16 support for Llama-v2 on \u2026", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 29, "deletions": 2}}, {"node": {"oid": "18f76bd25ded7a6ec4b8675e1c2813753fec5343", "messageHeadline": "Bump gradle/wrapper-validation-action from 1 to 2 (#19412)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f048fb5b14f5495950fb984dc474c8930861e474", "messageHeadline": "Bump nuget/setup-nuget from 1 to 2 (#19411)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1e10cdb2b9b9aa8ae6b248bca5ea139185e92947", "messageHeadline": "Fix subgraph quantization regression in onnxruntime 1.17 (#19421)", "author": {"name": "fxmarty", "email": "9808326+fxmarty@users.noreply.github.com", "user": {"login": "fxmarty"}}, "additions": 72, "deletions": 2}}, {"node": {"oid": "5c7e6b2e2af922e3dc7e58c584e454074a394d3d", "messageHeadline": "[EP Perf] Add CI option to enable TRT-OSS parser (#19448)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 63, "deletions": 29}}, {"node": {"oid": "5e70c6b3a68b7826e0fe8b3ebf8185ea59ace6de", "messageHeadline": "allow protobuf lite build for TRT EP (#19498)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 10, "deletions": 5}}, {"node": {"oid": "4dfba53bfb54e8fa50af52a536920d62840020a4", "messageHeadline": "[QNN EP] Build x64 python wheel for QNN EP (#19499)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 194, "deletions": 0}}, {"node": {"oid": "61e07a46e17c6dbc1adbd824d6b36aaac6f26c7b", "messageHeadline": "[DML EP] Support split hidden size for RotaryEmbedding (#18852)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 26, "deletions": 10}}, {"node": {"oid": "a622710fe171fab399bee41c8c3cd1dc16dd8f62", "messageHeadline": "Add option to skip session run in perf_test tool (#19501)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 20, "deletions": 1}}, {"node": {"oid": "7fa6f4fca4a05ac025995da70b7a5fa92ee46d83", "messageHeadline": "add arm64 bfloat16 fastmath mode option for transformers benchmarking\u2026", "author": {"name": "snadampal", "email": "87143774+snadampal@users.noreply.github.com", "user": {"login": "snadampal"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "90e2e8561f86efa634a7fe364d87e2d1dc159a69", "messageHeadline": "Ovep 1.17.1  (#19482)", "author": {"name": "Preetha Veeramalai", "email": "preetha.veeramalai@intel.com", "user": {"login": "preetha-intel"}}, "additions": 20, "deletions": 15}}, {"node": {"oid": "9cb97ee507b9b45d4a896f663590083e7e7568ac", "messageHeadline": "Disable CPU EP's allocator's arena when address sanitizer is enabled \u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 13}}, {"node": {"oid": "c831031ad54657916f0b0867327f11eb1691ee14", "messageHeadline": "Remove cuda gencode 90 to reduce onnxruntime-training package size (#\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d00adb7989635a4046d896fab6358ad4c7b695db", "messageHeadline": "Align bins_space_ storage (#17552)", "author": {"name": "Fangrui Song", "email": "i@maskray.me", "user": {"login": "MaskRay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1182b5509ba2604856d02cf22795d6874252892e", "messageHeadline": "Disable streams for the DML EP (#19481)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 76, "deletions": 11}}, {"node": {"oid": "0e984ef0d1c41cbc54a453f6de0c5e8784df5e74", "messageHeadline": "Small fixes in Resize CPU antialias (#19476)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "90cf03767dff037fb5cca3b45f51a2f6b021ac23", "messageHeadline": "Support ONNX export of OpenAi Whisper model (#17316)", "author": {"name": "Shubham Bhokare", "email": "32080845+shubhambhokare1@users.noreply.github.com", "user": {"login": "shubhambhokare1"}}, "additions": 474, "deletions": 29}}, {"node": {"oid": "1007d8f3d1904ff2efc4e0647e795939fe049464", "messageHeadline": "Revert \"Revert NeuralSpeed code for x64 MatMulNBits (#19382)\" (#19474)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1023, "deletions": 0}}, {"node": {"oid": "3d2ddf96e314ad97ee4377a730c465ffff4c8723", "messageHeadline": "Bump ruff linter to 0.2.1 (#19471)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 118, "deletions": 198}}, {"node": {"oid": "03be65e064c6cc2ae5a2169b13dcd675c1dc7cf8", "messageHeadline": "[js/web] fix types exports in package.json (#19458)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "148f54c6ea8b90dfc1dbf82197ac79c2102054b4", "messageHeadline": "Add capturestate / rundown ETW support logging for session and provid\u2026", "author": {"name": "ivberg", "email": "ivberg@microsoft.com", "user": {"login": "ivberg"}}, "additions": 164, "deletions": 26}}, {"node": {"oid": "3b1b18347ceb74dec8878fcf689e298ad9bc9d95", "messageHeadline": "Check for invalid combination of python + minimal build in build.py (\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "19952c5b3521934b8a34b6ef496cf3faa0c4eb59", "messageHeadline": "Add script to convert phi2 to ort-vllm compatible (#19429)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 183, "deletions": 43}}, {"node": {"oid": "0d10c7f3c1111cfff064e7990aa897ac9fd05c82", "messageHeadline": "Revert NeuralSpeed code for x64 MatMulNBits (#19382)", "author": {"name": "luoyu-intel", "email": "yu.luo@intel.com", "user": {"login": "luoyu-intel"}}, "additions": 0, "deletions": 1023}}, {"node": {"oid": "75f06319d61fd6c93387aa0f49b2ab7b2de647bf", "messageHeadline": "Change binet to bin (#19424)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "36d223676b419a8e6868d38f9363e236edbd0690", "messageHeadline": "Use GraphViewer.IsConstantInitializer in NNAPI EP. (#19401)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 382, "deletions": 346}}, {"node": {"oid": "302d4be7d935a6679a43ced5b3524b02f0e02da1", "messageHeadline": "[DML EP] Fix external data unpacking (#19415)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 20, "deletions": 8}}, {"node": {"oid": "91b2e660fe4b5ad70a7b6fde181cbfc9336e38a1", "messageHeadline": "[Build] fix: missing nvcc flags when compiling with unittests (#19308)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "df5c6718bd67bdc59fb77fbd4a6ccdcaf902145c", "messageHeadline": "Remove iOS simulator max runtime version limit. (#19396)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "bedf0eee737f1a8ff21de452887d233308446596", "messageHeadline": "[CUDA] Add use_tf32 provider option (for FP32 GEMM) (#19357)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 245, "deletions": 139}}, {"node": {"oid": "c4b49fb7bf340a0d27c7d8e2cb2508cac7f57ccf", "messageHeadline": "[CUDA] remove CUBLAS_TENSOR_OP_MATH mode (#19431)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 7, "deletions": 10}}, {"node": {"oid": "61b0e04b03ef2ce8b56b5a3a62ea1ae58e52ac73", "messageHeadline": "Bump github/issue-labeler from 3.3 to 3.4 (#19410)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cfe93a1eab4af4c911239ad13d7aeaa1307958b9", "messageHeadline": "fix type annotation for sess_options (#19393)", "author": {"name": "Yongxin Wang", "email": "fefe82@qq.com", "user": {"login": "fefe982"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5ff27ef02a1b8d8668c6a9f4da2b7a578f4d9f05", "messageHeadline": "[js/webgpu] support customop FastGelu (#19392)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 353, "deletions": 8}}, {"node": {"oid": "a4cfdc1c28ac95ec6fd0667e856b6a6b8dd1020c", "messageHeadline": "update comments for nodejs binding artifact preparation. (#19425)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 44, "deletions": 0}}, {"node": {"oid": "23996bbbbe0406a5c8edbf6b7dbd71e5780d3f4b", "messageHeadline": "[CUDA][ROCm][Training] Fix cuda/rocm provider info hash (#19398)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 75, "deletions": 38}}, {"node": {"oid": "06a84c8a0d738cf6c45a173e7283aa7dd1b6c2ba", "messageHeadline": "Enable DML on Windows and CUDA on Linux for Node.js binding (#19274)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 23, "deletions": 41}}, {"node": {"oid": "d2d9b5b5f9639a419412fc68174441330521af91", "messageHeadline": "fix output shape inference packed gqa (#19374)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "d120104dcd1e0942ca0856a096c55be05483d1d6", "messageHeadline": "add ATen support for bicubic interpolation (#19380)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 48, "deletions": 1}}, {"node": {"oid": "aaf32fb1b1d1cb3a3a4250cc2ee132576b7ddf00", "messageHeadline": "phi2 conversion/optimization script (#19338)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1801, "deletions": 3}}, {"node": {"oid": "e6d3518db9e6a19dc8088d5b9d5589a2b3a395d8", "messageHeadline": "Bump gradle/gradle-build-action from 2 to 3 (#19297)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "435e19953ea54115124fd637a67a87681a7fc8eb", "messageHeadline": "Fix llama.covert_onnx to make it runnable in CI (#19372)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 84, "deletions": 56}}, {"node": {"oid": "0cba56e0a010fe8c87a7be91bb84d36508edbf6d", "messageHeadline": "[ROCm] Fix CI pipeline by fixing pytest version (#19407)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "debd1cab10fd82394a416711890dc15fddec88dc", "messageHeadline": "Add coremltools 7.1 as a dependency (#19389)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 297, "deletions": 9012}}, {"node": {"oid": "18c3acb1983e402248365b9f2babfaf8e2182e81", "messageHeadline": "update import in convert_generation.py (#19385)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "ccbe264a39cf05acaebfa326edc9b9039c1771fe", "messageHeadline": "[js/webgpu] Add LeakyRelu activation for fusedConv (#19369)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 184, "deletions": 25}}, {"node": {"oid": "50806a7dd515a5378043117eb462362e773f6a8b", "messageHeadline": "[js/web] support external data in npm test (#19377)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 55, "deletions": 6}}, {"node": {"oid": "efc17e79de8c1a62eb419d19576ccb90b371b0d0", "messageHeadline": "[js/webgpu] Fix the undefined push error (#19366)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "9139bdda02d555417dcd5a97af411225bd35021b", "messageHeadline": "[ROCm] CK implementation support causal mask  (#18943)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 364, "deletions": 113}}, {"node": {"oid": "a2eb967008c0a9443d267b41c387e0600202fb3a", "messageHeadline": "Fix Split index bugs uncovered by QNN SDK 2.19 (#19381)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 61, "deletions": 20}}, {"node": {"oid": "09d5c1b56f94c9df216ef7de020b920fc52a0c14", "messageHeadline": "Fix DEBUG_GENERATION build (#19383)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 0, "deletions": 40}}, {"node": {"oid": "b71be3c1e31abba44f95ad6bc568ab0499b34eaa", "messageHeadline": "[VitisAI] Resolving compilation errors when using USE_VITISAI (#19368)", "author": {"name": "zz002", "email": "zhenze.wang@amd.com", "user": {"login": "zz002"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "3a2ab1963a195fe8df59e6220d3f191e5dfe80ee", "messageHeadline": "[js/webgpu] Refactor createTensorShapeVariables (#18883)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 40, "deletions": 64}}, {"node": {"oid": "13ad922e7ffd4eeeff5ca7199c4d5e7bf9703849", "messageHeadline": "Improve MatMulNBits test (#19378)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 6}}, {"node": {"oid": "8a2646ce604f7178ee5c57de5d66e7cca7356eb2", "messageHeadline": "Metrics - llama-2 - Add package name and version to engine of onnxrun\u2026", "author": {"name": "ironman", "email": "bitzhangxi@outlook.com", "user": {"login": "ironmangithub"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1bdd7d949905e487e15b8e3bc30d3cd197313bdd", "messageHeadline": "Update oneDNN to v3.0.1 in order to support gcc 13 (#19344)", "author": {"name": "He Li", "email": "li@imlihe.com", "user": {"login": "lihe07"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "319481898c6db9b1268ed23e9480bdfeb0b53bf5", "messageHeadline": "Give a triton library missing warning instead of silently turn off (#\u2026", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "0fa88bc8106acae37aa1876065e16fbf320b5b01", "messageHeadline": "Multi-partition support for context binary cache feature (#18865)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 919, "deletions": 764}}, {"node": {"oid": "eb0ce86db8eaa5a207dbd433ba8d0630dc60db4c", "messageHeadline": "[DML] Resize 18 & 19 (#19071)", "author": {"name": "Linnea May", "email": "linnealovespie@gmail.com", "user": {"login": "linnealovespie"}}, "additions": 242, "deletions": 62}}, {"node": {"oid": "1d6f13fb92204df344301ccc12a5f292a0cc44ed", "messageHeadline": "[VitisAI] Refactor the VAIEP to use MSFT's standalone API (#19058)", "author": {"name": "Yueqing Zhang", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 1000, "deletions": 1101}}, {"node": {"oid": "68b6064be66cc16bae7e624a554e641cf21d2b06", "messageHeadline": "Fix reporting of unused initializers in subgraphs (#19341)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "55b60d8fe04408f7326704f87b48c9ba318091e6", "messageHeadline": "Turn off Neural Speed to avoid slowdowns (#19265)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ca8d4459d44c55817b61178b4b573eb645b6c4ec", "messageHeadline": "Add contrib Q/DQ ops to symbolic shape inference tool (#19340)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 229, "deletions": 0}}, {"node": {"oid": "2b361c04d68ec21bdd1ede33bcdf7a142bd292d1", "messageHeadline": "Fix Flatbuffer build issue. (#19296)", "author": {"name": "Phoebe Chen", "email": "phoebe.chen@sifive.com", "user": {"login": "phoebesv"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "d87f73ab44a2a038f396cc0a38a7062c4516c59c", "messageHeadline": "[WebNN EP] Use GetVecUint32FromVecInt64 to simplify the code (#19324)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 7, "deletions": 23}}, {"node": {"oid": "3262e8df2f57c649ff0a75ccdc92e07ddaa8e4c4", "messageHeadline": "Introduce a Nominal Checkpoint for On-Device Training (#19232)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 973, "deletions": 311}}, {"node": {"oid": "4562c910fed0d8446497388415d103be40b65157", "messageHeadline": "Whisper Crash Fix (#19345)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "dd1f6ccc45e2cc852a5ff496019ce5d532855d76", "messageHeadline": "[js/webgpu] resolve codescan alert (#19343)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 5, "deletions": 7}}, {"node": {"oid": "d73131cf0f4eecd5f639c40d8fb6fad4efeaf4ef", "messageHeadline": "[js/webgpu] Use DataType as uniform cpu type (#19281)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 148, "deletions": 108}}, {"node": {"oid": "85cef0af8cdd99687402fe631bdba3d63926de6b", "messageHeadline": "[js/webgpu] Support capture and replay for jsep (#18989)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 436, "deletions": 136}}, {"node": {"oid": "6dd0079d133da09c286cb097b791f04ef387f5b2", "messageHeadline": "Exclude more code from custom_ops.cc when not required in minimal bui\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 508, "deletions": 450}}, {"node": {"oid": "1e936bfd6391b50a6cfc96730addd68ac398ef5e", "messageHeadline": "[WebNN] Ignore empty optional input tensor (#19235)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 40, "deletions": 23}}, {"node": {"oid": "e74f141338547d7eea9c6cbe23d1c892174163cf", "messageHeadline": "Save stablediffusion and open-clip in pipeline cache (#19314)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 64, "deletions": 16}}, {"node": {"oid": "0c38e96bb558295d68fadf33b6f835e4a181b8a0", "messageHeadline": "[Quant tool] Ensure MSFT opset for Q/DQ models (#19335)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "90883a366ae2ce5402c8b887638a8b2ae3e0efdd", "messageHeadline": "[js/webgpu] Add hardSigmoid activation for fusedConv (#19233)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 207, "deletions": 57}}, {"node": {"oid": "8dad9d92f4010b10f2d9cc08a5dca2929e5e8c68", "messageHeadline": "Move einsum's test data to constexpr variables (#19320)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1316, "deletions": 357}}, {"node": {"oid": "c379a89bcb26bc4838efe70fae04760106e8d081", "messageHeadline": "[MLAS AArch64] SQNBitGemm optimization (#19272)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 558, "deletions": 232}}, {"node": {"oid": "04afe77305c06181e31b8934df9ee8d3c19af2a7", "messageHeadline": "Update ThirdPartyNotices.txt: Add Intel neural-speed (#19332)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 207, "deletions": 0}}, {"node": {"oid": "febec1c5860c9b39e7ddd7167ea3cfa28ec2d2db", "messageHeadline": "Update Whisper export with beam search (#19322)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 18, "deletions": 5}}, {"node": {"oid": "3454f86e706a6df3bac52d105a338cb63674401f", "messageHeadline": "Windows - Only set thread affinity on Server with auto affinity (#19318)", "author": {"name": "ivberg", "email": "ivberg@microsoft.com", "user": {"login": "ivberg"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "b84cb247e3ef06639925120a84838ab970ef6843", "messageHeadline": "io_binding to handle optional input of sequence type_proto (#19273)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "ffc3431a660ba2fe3fb220be24f0ff3260d828bd", "messageHeadline": "Update ScatterElements to Support Opset 13, 15, 18  (#19198)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 858, "deletions": 25}}, {"node": {"oid": "3e17ca3dabd76d370827ef119f092be1b85422ea", "messageHeadline": "Fix iOS artifacts issue in Microsoft.ML.OnnxRuntime Nuget Package (#1\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 58, "deletions": 14}}, {"node": {"oid": "a92802f9403e3ca7313e7d29f663038669bffc57", "messageHeadline": "Disable a few tests for wasm build (#19316)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "9f68a27c7a3a932a574d50db19f40393a0cedf81", "messageHeadline": "[ORTModule] Handle Cast on Constant Number on Triton Code-gen (#19321)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 37, "deletions": 2}}, {"node": {"oid": "624b4e20635d37458a361a87e562f5ba5011f6cb", "messageHeadline": "[js/webgpu] Remove enableShapesUniforms (#19279)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 68, "deletions": 134}}, {"node": {"oid": "00d048121bfd99644ac2241e0d001d5394a970b6", "messageHeadline": "[TensorRT EP] Fix InferenceSession::Run() not thread-safe issue (#19301)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 43, "deletions": 11}}, {"node": {"oid": "465540d29b87bcbe5b4bc1d7ff4e7fb082693746", "messageHeadline": "Update training api python documentation (#19287)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 21, "deletions": 1}}, {"node": {"oid": "e91d91ae4f0057bd8ffcc3bd8c5826af85f77731", "messageHeadline": "Fix a build issue: /MP was not enabled correctly (#19190)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 72, "deletions": 69}}, {"node": {"oid": "4ee222413f4ac0bef1d7383c693796b9b5a30106", "messageHeadline": "Update OneBranch.Nuget-WindowsAI-Pipeline.Official.yml for Azure Pipe\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "9e69606360d7e77f9ee869beec2b8c9e43517fae", "messageHeadline": "fix f16 for attention, enable slice and flatten for more types (#19262)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 9, "deletions": 13}}, {"node": {"oid": "e96a038f01ab198cb03dea3b918351e3cfe4a9fc", "messageHeadline": "Add VP test in Stable diffusion pipeline (#19300)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 131, "deletions": 6}}, {"node": {"oid": "82c1cb416b8054f67fe1f73928ad4c276d80afdb", "messageHeadline": "[CUDA] Refactor GroupNorm and add common vectorize implementation (#1\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 217, "deletions": 102}}, {"node": {"oid": "6d7ac9c93ae7cd1f979b35e7f6d7af207962cd99", "messageHeadline": "Support general session config entries in perf test tool (#19289)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 78, "deletions": 4}}, {"node": {"oid": "d7ff81dfb77989a8ce975db29457e5cdfc00f9e3", "messageHeadline": "[CUDA] support user_compute_stream in python API (#19229)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 35, "deletions": 0}}, {"node": {"oid": "7d4dc66846aadb2daf63fea3504aff0c596d1d38", "messageHeadline": "ExecutionProvider API refactor - make GenerateMetaDefId a standalone \u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 187, "deletions": 147}}, {"node": {"oid": "fc44f96ad523526b23d5e6851bd89f888e0de2bc", "messageHeadline": "Add support for a collection of OrtValue as inputs and outputs to C# \u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 182, "deletions": 0}}, {"node": {"oid": "358650d4415d930ba3ea4de159b8191cb1696dc4", "messageHeadline": "Fix BigModel stable diffusion pipeline (#19277)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3, "deletions": 8}}, {"node": {"oid": "a3f0e2422b5eb2968e3f11e93414aa1661b32e2f", "messageHeadline": "[js/webgpu] Support f16 uniform (#19098)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 56, "deletions": 26}}, {"node": {"oid": "8b4517218b52285efaaf8badd303c00b0e514238", "messageHeadline": "Remove USE_CUTLASS flag (#19271)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 25, "deletions": 131}}, {"node": {"oid": "656ca66186c7fd362abd8f33915bd0f96483bf43", "messageHeadline": "[js/webgpu] Support uniforms for conv, conv transpose, conv grouped (\u2026", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 418, "deletions": 344}}, {"node": {"oid": "a2867b911e67146218b4fc0b32721e5cdbade49b", "messageHeadline": "[TensorRT EP] Fix mem leak for TRT plugins custom ops (#19248)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 35, "deletions": 75}}, {"node": {"oid": "2b285cd78a629971a9e465036e94a431e6fef17b", "messageHeadline": "[CUDA] Add functions to dump bfloat16 tensors (#19266)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 75, "deletions": 40}}, {"node": {"oid": "5b065050734e6bc397dc38ba0df246aeb57ac508", "messageHeadline": "[js/webgpu] Fix Tanh explosion (#19201)", "author": {"name": "Jiajie Hu", "email": "jiajie.hu@intel.com", "user": {"login": "hujiajie"}}, "additions": 30, "deletions": 1}}, {"node": {"oid": "1c92e56dc0f906a43128e2f0c4c6729349aac92b", "messageHeadline": "[Cuda] Refactor GroupNorm (#19146)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 653, "deletions": 521}}, {"node": {"oid": "2b87dd373a3567c2c426e2f090b201b8b051a346", "messageHeadline": "[ORTModule] Remove Mod from Hash to Avoid Conflict for Triton Code-ge\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "7dd1f4b8e27f38b55f2430f84ddaae1128bef9f4", "messageHeadline": "Pad-18 Cuda implementation (#19211)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 287, "deletions": 171}}, {"node": {"oid": "4477f57ee3151287a9759bd09d269f0e258a9eda", "messageHeadline": "Enable RISC-V 64-bit Cross-Compiling Support for ONNX Runtime on Linu\u2026", "author": {"name": "Phoebe Chen", "email": "chen.miat@gmail.com", "user": {"login": "phoebesv"}}, "additions": 206, "deletions": 3}}, {"node": {"oid": "0c2f0ba90da11ad53c63810e5f3e6fda4e295899", "messageHeadline": "[WebNN EP] Support conv1d by reshaping with prepended 1's (#18857)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 141, "deletions": 89}}, {"node": {"oid": "7252c6e747de83b65285601281a9d07aea801fba", "messageHeadline": "[WebNN EP] Support WebNN async API with Asyncify (#19145)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 30, "deletions": 49}}, {"node": {"oid": "c456f19dbaf6b23928a60e8b356a429ae76376a4", "messageHeadline": "remove old quantization tool file (#19247)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 0, "deletions": 414}}, {"node": {"oid": "591f90c0b9e8d0922fcebabffed8d54b67d7a613", "messageHeadline": "[js/webgpu] Fix issue of timestamp query (#19258)", "author": {"name": "Yang Gu", "email": "yang.gu@intel.com", "user": {"login": "gyagp"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "bc54ad3f03d7ee333f5e0c62ebf892c32f8a51a5", "messageHeadline": "Update abseil to a release tag and register neural_speed (#19255)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 24, "deletions": 16}}, {"node": {"oid": "a28abeb24100441c76a777f9ce225cb0ea3a59c3", "messageHeadline": "Change \"#ifdef WIN32\" to \"#ifdef _WIN32\" (#19254)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "a33b5bd1fa5ac6d9aabb23cd8aca16b5fc3fc3c5", "messageHeadline": "[JS/WebGPU] Added Uniforms to SkipLayerNorm. (#18788)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 69, "deletions": 58}}, {"node": {"oid": "a39ac4a97976c9bea49be6e646ac1fac64278f65", "messageHeadline": "[DirectML] Register Pad19 (#19175)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 16, "deletions": 0}}, {"node": {"oid": "d7aebf9ea8a4a651088384f219292bae9062439b", "messageHeadline": "Move Nuget Test from T4 to A10 to reduce release duration (#19253)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "c10be1848cafa7575ba298cbcc01e89dcd841851", "messageHeadline": "[TensorRT EP] Avoid calling unavailable function with cpu python pack\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "6a424ccf8c2f9cd7f191c843547d5f37ef409493", "messageHeadline": "Fix AMD pipeline test failures (#19250)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "cbb29d80ff5ec63d3cc2289911c4420f5a9d8a2d", "messageHeadline": "GQA Rotary and Packed QKV with Flash (#18906)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 1517, "deletions": 272}}, {"node": {"oid": "532f8c642ce9c1ea2971b7d0f0ff8a4197bcb3a0", "messageHeadline": "Fix a backend test by using local backend (#19230)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "f53068446e7e560012862e1812270bcf908fbda4", "messageHeadline": "Add Temperature to WhisperBeamSearch input (#19188)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 20, "deletions": 4}}, {"node": {"oid": "54871a27736cf54cbda9c4f09bb27e931de7334e", "messageHeadline": "Replace T4 to A10 in Linux GPU workflow (#19205)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 31, "deletions": 4}}, {"node": {"oid": "0ea48fc73ec6bdbb8af2010483a61823fcf1a613", "messageHeadline": "Modified the condition to load the optimiser model (#18891)", "author": {"name": "Heflin Stephen Raj", "email": "heflinstephen03@gmail.com", "user": {"login": "heflinstephenraj"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "61610ff9862ad834f153ed3e70ba526dac86ae7c", "messageHeadline": "[js/webgpu] Add FusedConv clip test case (#18900)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 34, "deletions": 0}}, {"node": {"oid": "6ca7c1a933e57e0078d8d01eff3a1520098cfed1", "messageHeadline": "unet fusion for stable diffusion webui (#19227)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 395, "deletions": 82}}, {"node": {"oid": "b2aec41a8309bc2dced74a991b1f3c311e037e3d", "messageHeadline": "[ROCm] enable hipGraph (#18382)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 241, "deletions": 42}}, {"node": {"oid": "37d14d78960fb1ba54c0bb2dc3be740e93d2ca15", "messageHeadline": "[QNN EP] Create Windows ARM64 nightly python package (#19128)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 185, "deletions": 1}}, {"node": {"oid": "d226e40856738531cf8b481b07379545f7cfefe2", "messageHeadline": "[js/webgpu] set query type in onRunStart (#19202)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 13, "deletions": 5}}, {"node": {"oid": "2e0a388c36b92bc412dfa8ad45af23c7f28a4d49", "messageHeadline": "[js/webgpu] Add HardSigmoid support (#19215)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 30, "deletions": 3}}, {"node": {"oid": "e283cdb21857170848e9b8a8fbca24d0463b4193", "messageHeadline": "Fix Fuzz Testing CI (#19228)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "24b74aebcbd5fbaaa44ca41143b3b6afe3207978", "messageHeadline": "[DML] Register DML operators for opset 19  (#16939)", "author": {"name": "Linnea May", "email": "linnealovespie@gmail.com", "user": {"login": "linnealovespie"}}, "additions": 59, "deletions": 32}}, {"node": {"oid": "77da2ef278a4e77cca4cef4e5d72ed1ef46fcce3", "messageHeadline": "[aarch64] Add Sbgemm kernel to accelerate fp32 tensor matmul with bfl\u2026", "author": {"name": "snadampal", "email": "87143774+snadampal@users.noreply.github.com", "user": {"login": "snadampal"}}, "additions": 3473, "deletions": 18}}, {"node": {"oid": "780acda7b4f044564e1f222901fd6a676aa05cbf", "messageHeadline": "Add Big models pipeline (#19222)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 259, "deletions": 0}}, {"node": {"oid": "8d9d7511799e2138c14454bb672caf07dcdc2457", "messageHeadline": "[QNN EP] Expose device-level session options (#19212)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 292, "deletions": 118}}, {"node": {"oid": "373ebac167a1d7f1dfb7a576a6c92f1c75cb711e", "messageHeadline": "Zhalei/fix seqoutput type (#18765)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 220, "deletions": 69}}, {"node": {"oid": "21034a2c37d707ee913dcca1b00d0b9e7651f980", "messageHeadline": "phi2 contrib ops changes (#19112)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 280, "deletions": 81}}, {"node": {"oid": "f3402de01e732283283aaa208022d6c7ae85ca4a", "messageHeadline": "[TensorRT EP] Enhance EP context configs in session options and provi\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 624, "deletions": 162}}, {"node": {"oid": "c8ce83967e5b52062558046d769f3af7d871e893", "messageHeadline": "Download protoc for all Apple host builds, remove protoc build from i\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 40, "deletions": 41}}, {"node": {"oid": "6e17571f2f0cb8eb841395f208d8c31359e2f054", "messageHeadline": "Fix issue that the generated context cache model inputs/outputs order\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 210, "deletions": 26}}, {"node": {"oid": "a3ecb6326711ca368ccc402886dcaf1ef56ff79c", "messageHeadline": "Update LLaMA attention fusions (#19200)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 46, "deletions": 25}}, {"node": {"oid": "eaf047c82092062e5a5df7f4087a9a5c7ec75f2a", "messageHeadline": "Increment year to 2024 in conf.py (python documentation) (#19107)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "28a16c223cb7a8707fb90340c8aecdc46eec3f2f", "messageHeadline": "[QNN EP] Update QNN pipelines to use QNN SDK 2.18 by default (#19129)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 11, "deletions": 23}}, {"node": {"oid": "d69b622ef489dfa9e9fdd56f6ce6df7376f9fa6f", "messageHeadline": "[js/web] upgrade dependency packages version (#19193)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 34, "deletions": 48}}, {"node": {"oid": "dc1fed7268876bcf8c12161d9398a65c427af315", "messageHeadline": "[Fix] Dual Cuda version isn't supported as expected in Linux Gpu pipe\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 52, "deletions": 45}}, {"node": {"oid": "459c750b031339456e4061b1c4214904e6853ccd", "messageHeadline": "Update x64 template kernel library for 'sqnbitgemm' (#19016)", "author": {"name": "luoyu-intel", "email": "yu.luo@intel.com", "user": {"login": "luoyu-intel"}}, "additions": 753, "deletions": 24620}}, {"node": {"oid": "dd2177c5d70b8e5b704f7ee0ddce134243eacb24", "messageHeadline": "enable webnn in ci build (#19163)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "dadd3ea704243a8c2b2ded790ae01f3b57c4da53", "messageHeadline": "Check the ep_cache_context and don't allow access outside the directo\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 155, "deletions": 2}}, {"node": {"oid": "9da3e36138dd24377fbb0b4022d891b3baf07b84", "messageHeadline": "Fix buildJava from Zip-Nuget-Java-Nodejs Packaging Pipeline (#19187)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "f87e69801f200a34ddb312f1d39e7296f19b660b", "messageHeadline": "[js/web] show warning when numThreads is set but threads is not suppo\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 33, "deletions": 6}}, {"node": {"oid": "146ebaf91e85185a0ac18c82bc69eba685ab9727", "messageHeadline": "[js/web] allow proxy to load model with 1GB <= size < 2GB (#19178)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 13, "deletions": 3}}, {"node": {"oid": "bc219ed553fc8d4b8fa3c7b4476810a63a864d8b", "messageHeadline": "[TensorRT EP] Enable a minimal CUDA EP compilation without kernels  (\u2026", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 97, "deletions": 39}}, {"node": {"oid": "bd9d8fb2a545a59d87a4c23308ec543ba6e4c41d", "messageHeadline": "[ORT 1.17.0 release] Bump up version to 1.18.0 (#19170)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 31, "deletions": 26}}, {"node": {"oid": "63dd605d3310f5a9540c414216f3f3b67d455c4a", "messageHeadline": "Fix untyped float values in quantization tool missing from PR #18043 \u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 131, "deletions": 23}}, {"node": {"oid": "9876cc7c4f5f6249e1dec8b93abf7b8dfcf5ca0c", "messageHeadline": "more inputs support for LLM exporter (#19005)", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 16, "deletions": 5}}, {"node": {"oid": "07d3aed3aa3a054deb502cedf867f559fc690755", "messageHeadline": "[WebNN EP] Fixed build issue with disable_rtti (#19173)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "81d363045ba273b16a3ec654c53a15217a2d2a36", "messageHeadline": "Upgrade Ubuntu machine pool from 20.04 to 22.04 (#19117)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 50, "deletions": 60}}, {"node": {"oid": "e61861b0a121bca1d60e5d4a3722e52b6820c430", "messageHeadline": "Clean up generated files in QNN UTs (#19127)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "c935c8fbd2e463a3e0153145140a8efd780dfabc", "messageHeadline": "remove unnecessary environment variable (#19166)", "author": {"name": "moyo1997", "email": "54333118+moyo1997@users.noreply.github.com", "user": {"login": "moyo1997"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "8e272b9cac70a11c472fb002af755213a4dabf66", "messageHeadline": "Update build.py to remove unused functions and update python to 3.8 (\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 31}}, {"node": {"oid": "80f274ca6f2f4572d827edd6dc7f736d7a8c036a", "messageHeadline": "Fix SkipLayerNormalization shape inference (#18724)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 43, "deletions": 3}}, {"node": {"oid": "e2e488d6f8bcd14f40e9e2c8e65f310ce9c0e872", "messageHeadline": "Revert \"iOS packaging pipeline stability\" (#19135)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 42, "deletions": 41}}, {"node": {"oid": "c92f72ebebf5f4a1e63b726e6e5cec1a47250bb5", "messageHeadline": "Merge Linux Nuget GPU pipeline with zip-nuget (#19120)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 31, "deletions": 161}}, {"node": {"oid": "8d4369b77ef8567653db3e247bbb2f48889fc457", "messageHeadline": "Update DirectML nuget version to 1.13.1 (#19122)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "1bab98988b4e7b6d33be0e672fce361ccbb1d397", "messageHeadline": "[WebNN EP] Fixed bug in int8 data type processing (#19134)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 36, "deletions": 10}}, {"node": {"oid": "9dee543bedaed8419957afaed3a64b1ab5fa3a21", "messageHeadline": "fix gemm beta for fp16 (#19153)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9f87c5c41d50fdcf30ce439617c708c964d8a050", "messageHeadline": "Fix build error due to merge with DML adapter enumeration macro defin\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "1150b1f81ea7e46a840212acf194422af7f764a3", "messageHeadline": "ORTModule memory improvement (#18924)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 638, "deletions": 30}}, {"node": {"oid": "191525301f2b30fa4ff7337cd40c5f3f94834488", "messageHeadline": "[java] Updating TensorInfo so it contains the named dimensions (#18962)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 83, "deletions": 12}}, {"node": {"oid": "a97199c62de4a96939624ba511313d0f81014f56", "messageHeadline": "Fix Arm64EC build for test_q4qdq.cpp (#18523)", "author": {"name": "Ben Niu", "email": "niuben003@gmail.com", "user": {"login": "mcfi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "922a2f00e3855fdc9852ed1bfe7f6f0a88e40a24", "messageHeadline": "Extend timeout in Nuget-CUDA-Packaging-Pipeline (#19138)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b2ce3eedb9f3d9cee82525c9f29c2d1f42ba58c7", "messageHeadline": "Fix build error for CoreML Split op (#19099)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "71657d1eb8b0a24a4b6584d9e904506a0b4e1521", "messageHeadline": "[java] Fix double close (#19133)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 208, "deletions": 18}}, {"node": {"oid": "c3ce9df80c2cfc7013445f8b44213f3e75cac753", "messageHeadline": "Disabling python3.12 on training python packaging pipleines (#19123)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "76797127d6a3125fc59e605670809957a2183cbe", "messageHeadline": "Always download cuda and trt libraries from Azure blob (#19118)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 23, "deletions": 15}}, {"node": {"oid": "bb4011b2b14cb2702a4922ccd0b070d9ecc49a93", "messageHeadline": "Set default flags nvcc and do not set default compile flags for ROCM \u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 18, "deletions": 2}}, {"node": {"oid": "f917dde71740982c4520febc0ced1bff58b0068d", "messageHeadline": "[web] remove xnnpack from web backends (#19116)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 14, "deletions": 21}}, {"node": {"oid": "e1e45901e2fe9e22f14bf03398d4212671dc37ce", "messageHeadline": "iOS packaging pipeline stability (#19097)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 41, "deletions": 42}}, {"node": {"oid": "5558912d7b3ed85f7b1648e78ee8e66f717bd28f", "messageHeadline": "Disable ccache in Windows CPU CI pipeline (#19131)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "65893ef3822e31f095d4f59563a2145e52e8f218", "messageHeadline": "Add --parallel to QNN EP NuGet pipeline build command (#19126)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e803f8eb0f6cac716c1c9c2a548867f3349c9ffc", "messageHeadline": "[js/webgpu] Refactor timestamp-query and introduce timestamp-query-in\u2026", "author": {"name": "Yang Gu", "email": "yang.gu@intel.com", "user": {"login": "gyagp"}}, "additions": 198, "deletions": 123}}, {"node": {"oid": "78e796bb278ab5789c4f868518b6df7da39d8477", "messageHeadline": "Fixing issue where unzip package froim 'onnxruntime-win-x64-gpu' was \u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 19, "deletions": 2}}, {"node": {"oid": "07cfc56538bf7aa769943dbf96647cb2b677deb8", "messageHeadline": "[js] enable external data loading for ort-web (#19087)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 420, "deletions": 108}}, {"node": {"oid": "e5eacc6d1192c1ced310e93013ce917e2fe023a7", "messageHeadline": "Fix cuda-packaging-pipeline.yml (#19115)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "62a4e9103e33b3372d262ad490a9d1c67d84b2eb", "messageHeadline": "Add extreme_power_saver for htp_performance_mode (#19111)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 25, "deletions": 10}}, {"node": {"oid": "443aeb851c7f941237ffecb04d36462216eaeaf3", "messageHeadline": "[TensorRT EP] Customizable engine cache prefix (#19083)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 120, "deletions": 8}}, {"node": {"oid": "150c4cb8fe005d8f0befc46239eac8770d4d5c68", "messageHeadline": "[MLAS AArch64] SQNBitGemm CompInt8 kernel (#18953)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1672, "deletions": 564}}, {"node": {"oid": "a756017e9f9a6a5d66a4b5b36888c275d84220d5", "messageHeadline": "[js/webgpu] more fixes for access above 2GB (#19065)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 67, "deletions": 71}}, {"node": {"oid": "8deeba3ad01ec2b7f815a620630659a1d2722f7f", "messageHeadline": "[Quantization] Fix get_qnn_qdq_config to use new scale/zp np.array da\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 148, "deletions": 86}}, {"node": {"oid": "96dbac6e4b996dfc5fcebd60d2f093beea87401e", "messageHeadline": "update to emsdk-3.1.51 (#18844)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 12, "deletions": 13}}, {"node": {"oid": "8f2e57f5d0dcbc4577816fcfee01fecdadc7e646", "messageHeadline": "Make session configuration options available to kernels via OpKernelI\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 246, "deletions": 162}}, {"node": {"oid": "a503561d0c7a0da644b4c6477b4c4f8da9524c88", "messageHeadline": "[js] using OffscreenCanvas when DOM is not available (#19033)", "author": {"name": "Jiangzhuo", "email": "jiangzhuo9357@gmail.com", "user": {"login": "jiangzhuo"}}, "additions": 37, "deletions": 12}}, {"node": {"oid": "4a5f13b681c4d8cc4d0bb58f2451342933a201a5", "messageHeadline": "fix resize for fp16 (#19110)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 15, "deletions": 12}}, {"node": {"oid": "4dbaa7373875d8366d1b47691c86386fee614fe2", "messageHeadline": "[js/web/training] added end-to-end tests (#18700)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 559, "deletions": 0}}, {"node": {"oid": "c340bf08f602ba269245c46244435810ee3a52c3", "messageHeadline": "Openvino EP code changes for 1.17 update (#19023)", "author": {"name": "Preetha Veeramalai", "email": "preetha.veeramalai@intel.com", "user": {"login": "preetha-intel"}}, "additions": 201, "deletions": 113}}, {"node": {"oid": "dcd6d4cad6ec7406cea192cbbbcda80c336b98c2", "messageHeadline": "Label encoder opset4 (#17977)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 1587, "deletions": 813}}, {"node": {"oid": "55b046e97e799e5d71901f9e51fea27e35c15a95", "messageHeadline": "Remove enable_mac_silicon settings (#19108)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "4520b76417200152ea3c8c68c5a685630c6b5304", "messageHeadline": "Exclude TP custom API from minimal (#19086)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 13, "deletions": 0}}, {"node": {"oid": "3c0a6b505a476c2915e89c0d0ebe8151105cf879", "messageHeadline": "Update transformers module to 4.36 (#18993)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3eec1592bd87deede7834cdca4b6d0e4baed14dc", "messageHeadline": "[WebNN EP] Update WebNN unit test list (#19103)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 205, "deletions": 206}}, {"node": {"oid": "c23410a182bab53fd90080887debd71290c4d84e", "messageHeadline": "StringSplit operator (#18016)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 248, "deletions": 7}}, {"node": {"oid": "2cb5781b439d12e23d8ccbdb63e8bc9b8bf7e2d4", "messageHeadline": "Remove two tests from test_logging_apis.cc (#19100)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 17}}, {"node": {"oid": "c8399a81fed9c114c43daf2103fee48d6b02bdd7", "messageHeadline": "Quantization tool: support float 8 with MatMul, support float 16 weig\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1107, "deletions": 237}}, {"node": {"oid": "0e8d4c3d2119f74df774420e4991bb3d9c09ed95", "messageHeadline": "Enable Address Sanitizer in CI (#19073)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 386, "deletions": 213}}, {"node": {"oid": "285606108ae7afae96e8a0cc7932bc5bf05640aa", "messageHeadline": "Set pythonInterpreter in set-python-manylinux-variables-step.yml (#19\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "e3ee255950a125f571343ad65037734f7b4feff5", "messageHeadline": "Remove the references to CreateFileMapping2 (#19102)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 13}}, {"node": {"oid": "e1db44b4f0e819db7eb85d0a0039bee23d762d40", "messageHeadline": "[WebNN EP] Add quantize Ops (#18011)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 232, "deletions": 4}}, {"node": {"oid": "acba63c36a69966ed90991ef9cba782326b6451e", "messageHeadline": "[js/webgpu] Change A/sqrt(B) to A*inverseSqrt(B) in normalization ops\u2026", "author": {"name": "Jiajie Hu", "email": "jiajie.hu@intel.com", "user": {"login": "hujiajie"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "5373c0c7300e607ffdc008a26c7bc4a97ec8f1f3", "messageHeadline": "Bump follow-redirects from 1.15.2 to 1.15.4 in /js/web (#19068)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "e2c145d37f6fb8a575c70798eaa03d7fedf9d23a", "messageHeadline": "Add Anubis metrics schema for local benchmark results uploading (#19018)", "author": {"name": "gunandrose4u", "email": "52735340+gunandrose4u@users.noreply.github.com", "user": {"login": "gunandrose4u"}}, "additions": 300, "deletions": 6}}, {"node": {"oid": "46dd0d3f52183d951640258c09c54adc1daeeeb9", "messageHeadline": "[TensorRT EP] Load precompiled TRT engine file directly  (#18217)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1873, "deletions": 840}}, {"node": {"oid": "b6d82834d4c311757d0444a9f3ae26e3ab0c551d", "messageHeadline": "add bfp16 to gqa (#19095)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 309, "deletions": 15}}, {"node": {"oid": "189be8e997ff84f30fca5aa9a070c139a4b337df", "messageHeadline": "Bump follow-redirects from 1.15.2 to 1.15.4 in /onnxruntime/test/wasm\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "d8962d67f48c957d7cbe15f86bcb70c0f7b6074b", "messageHeadline": "RegexFullMatch operator (#18002)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 177, "deletions": 3}}, {"node": {"oid": "08cf4fbcad784bd69a9e9d8102c1800ea9dbc2f1", "messageHeadline": "Handle all float types in IsQDQPairSupported (#19085)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 53, "deletions": 31}}, {"node": {"oid": "8a0a972f39f662e25567c455d9ff2b4b2ac66bec", "messageHeadline": "Update DML EP to accept broadcasted tensor of size 1 to match CPU (#1\u2026", "author": {"name": "Christian Larson", "email": "chrilaMSFT@users.noreply.github.com", "user": {"login": "chrilaMSFT"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "daa22f919feb0280e2951e5345483c3c8a8afebf", "messageHeadline": "[TensorRT] query GPU properties only once when setting device_id (#19\u2026", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 12, "deletions": 16}}, {"node": {"oid": "4d1243b4b4dd5ce9475a690cb270e7435f3e970b", "messageHeadline": "ORT ETW dynamic logging that improves ORT diagnosability & performanc\u2026", "author": {"name": "ivberg", "email": "ivberg@microsoft.com", "user": {"login": "ivberg"}}, "additions": 591, "deletions": 123}}, {"node": {"oid": "d0bac8216d324e78e6f3a89453bac207522b93eb", "messageHeadline": "[js/webgpu] fix bcast in where (#19009)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "53497702a6647e853af0b97a913010a5b09e4114", "messageHeadline": "Fix Nuget CUDA Packaging pipeline (#19054)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 52, "deletions": 9}}, {"node": {"oid": "24e9daf7071b4a37be92e1b6e5bbf5cbaa0e65b6", "messageHeadline": "Enrich cuda resources with ep options (#19014)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 91, "deletions": 48}}, {"node": {"oid": "58bf836592cbecdc53c197e9f1b45345b5e3ab3b", "messageHeadline": "Offline tooling for training to use reduction with keepdims=False (#1\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 40, "deletions": 8}}, {"node": {"oid": "4694edcd4199419ec9315cb94305f906c1df49ba", "messageHeadline": "String concat operator (#17994)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 156, "deletions": 5}}, {"node": {"oid": "f68dfcd8887ba1993143a0e496e127e7af2d4f4a", "messageHeadline": "[CUDA] Improve performance of DecoderMaskedMultiheadAttention on A100\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 253, "deletions": 52}}, {"node": {"oid": "2eb3db6bf03b3f190d20542bfdbd9dd0a84881c8", "messageHeadline": "Adding python3.12 support to ORT (#18814)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 103, "deletions": 10}}, {"node": {"oid": "a89db01fceafde24669f44b7f8f52a86b8ce63a6", "messageHeadline": "[js/webgpu] disable GroupedConvVectorize path (#19090)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "f11713702f8cfac5785c970d96c455e883d6c269", "messageHeadline": "Bump follow-redirects from 1.15.2 to 1.15.4 in /js/node (#19070)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "d03e477b9026a97d22dba64cd00b4614603671e5", "messageHeadline": "Fix missing subgraph candidates for recompute (#19077)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 83, "deletions": 41}}, {"node": {"oid": "0a0ef958eb94d94a17fd8c09a6d217e0589827a2", "messageHeadline": "update .vscode/settings.json (#19084)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "053ddfe3fd52135742567940750b7bf6ccffe166", "messageHeadline": "Disable per-session thread pool for web (#18480)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 28, "deletions": 1}}, {"node": {"oid": "5678317bafd219e2b71c72431905b776460e11a4", "messageHeadline": "Fix the duplicated QDQ attributes setup issue (#18039)", "author": {"name": "Yvonne Chen", "email": "yche@synopsys.com", "user": {"login": "vera121"}}, "additions": 41, "deletions": 1}}, {"node": {"oid": "fd6bab4250c41a7f6498e6fa02ba446bc74e0a8d", "messageHeadline": "[js/webgpu] Provide a vectorized algorithm for GroupedConv (#18884)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 271, "deletions": 6}}, {"node": {"oid": "e58319ebfc344419b94ab5f8f27f7ce5eabe56f5", "messageHeadline": "[TensorRT EP] Fix memleak (#19053)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "731b50dfc4f8074185dc70f3a10236fa4fdfc0aa", "messageHeadline": "Support INT4 weight only quantize, including RTN and GPTQ 2 algorithm\u2026", "author": {"name": "yuwenzho", "email": "yuwen.zhou@intel.com", "user": {"login": "yuwenzho"}}, "additions": 246, "deletions": 24}}, {"node": {"oid": "df116b82c743f9104bd090a310d5777a4475e539", "messageHeadline": "Custom op API for thread pool  (#18980)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 80, "deletions": 4}}, {"node": {"oid": "cf78d01546ca059a2ab487e01626e38029a3e8fd", "messageHeadline": "remove use of ai.onnx.ml in test for custom ops and local functions (\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "5f3113ecd612b4ef070d5a3c8a586d725cfdb13b", "messageHeadline": "[ROCm] Fix hipify error: fast_divmod.h: No such file or directory (#1\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "ed0f26d3d4d72d1a449f759ee27bebf128633911", "messageHeadline": "[js/webgpu] Revert parse norm attributes (#19074)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "730df1bfa284ffee7bc777de83b2c5e80c7cd98e", "messageHeadline": "Increase MacOS pipeline timeout (#19072)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "b25980c011f658709404f1b850302f94fc9541f4", "messageHeadline": "Disable rust pipeline for now (#19067)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 176}}, {"node": {"oid": "fa14dcd2b66d8e1f45a78a81f71276770f10aadf", "messageHeadline": "[WebNN EP] Support subgraph of the control flow nodes (#18923)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 62, "deletions": 6}}, {"node": {"oid": "76dfe5347cfdc00a555fb4e1867acf77dad5c819", "messageHeadline": "[js/webgpu] Support uniforms for instance-norm (#18929)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 249, "deletions": 86}}, {"node": {"oid": "37ac9d391ccd2bb3ea331d280541b675d650209b", "messageHeadline": "Enable Arm Compute Library 23.08 (#17672)", "author": {"name": "Milos Puzovic", "email": "79916358+milpuz01@users.noreply.github.com", "user": {"login": "milpuz01"}}, "additions": 113, "deletions": 20}}, {"node": {"oid": "a2afd92093fea3351b8dce80b142e3eb70dcf022", "messageHeadline": "Format TS code (#19066)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 5}}, {"node": {"oid": "897a4163d7638f680902114967e8888f42ee665b", "messageHeadline": "Update transformer version for training CIs (#19046)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "574c7caf3ac4c6b0c1004b1fd9e2b020c19d441b", "messageHeadline": "[TensorRT EP] Clear constrain of trt plugin with different input type\u2026", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "ad6dd0a597256d980a357751c347670c3bdc824f", "messageHeadline": "[WebNN] Enable npm unit tests (#18486)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 1072, "deletions": 94}}, {"node": {"oid": "557ac74c0565b7cfdb68bb79724d7ed1b870c121", "messageHeadline": "[js/webgpu] Support gemm uniforms (#19056)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 64, "deletions": 59}}, {"node": {"oid": "42ba2aed54458a2b645d859efb830a3c431ec896", "messageHeadline": "[js/webgpu] Support pad uniforms (#19057)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 107, "deletions": 119}}, {"node": {"oid": "eb92681bfbb95e93e34f301605529d605f7f9e63", "messageHeadline": "[js/webgpu] Support range uniforms (#19055)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 27, "deletions": 14}}, {"node": {"oid": "c1367ae5537162e8030a584d5d3e340d58a1bfe9", "messageHeadline": "Sqnbitgemm: add loongarch64 code path (#18775)", "author": {"name": "junchao-loongson", "email": "68935141+junchao-loongson@users.noreply.github.com", "user": {"login": "junchao-loongson"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dee6a5b3715c5bdf7a6d29c2b9516902ebd0e0b1", "messageHeadline": "[js/webgpu] Support uniforms for attention and multihead attention (#\u2026", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 190, "deletions": 186}}, {"node": {"oid": "ab897a4a4064e8259ed952cbd33c35d3ddd5f370", "messageHeadline": "Remove Windows ARM32 from nuget packaging pipelines (#19049)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 49, "deletions": 132}}, {"node": {"oid": "7cb8b20db2d329cf67e170293b2d2c81213e6100", "messageHeadline": "Remove mem consuming test case to unblock running ci on lower-end gpu\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 0, "deletions": 6}}, {"node": {"oid": "eb35896ede6e77bf4a453b9e7314e728e40f96ba", "messageHeadline": "[WebNN EP] Update WebNN normalization ops (#18817)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 57, "deletions": 92}}, {"node": {"oid": "68c29ece23821b1d2b73ac55c2a4266c72865219", "messageHeadline": "In a Linux or Android build check if the compiler support bfloat16 an\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 44, "deletions": 8}}, {"node": {"oid": "8f024b739439c521cd19fe5a9830d4015de99bd7", "messageHeadline": "[js/webgpu] Support uniforms for layer-norm (#18755)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 46, "deletions": 41}}, {"node": {"oid": "a8bb1df331e56e3a65106578b6475d89d17b27c5", "messageHeadline": "[js/webgpu] fix heap access > 2GB (#19010)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 47, "deletions": 46}}, {"node": {"oid": "975a315cd70aac54a6c4ff8b7d4e0a76d25666de", "messageHeadline": "Fix x86 build error in GraphDescBuilder.cpp affecting packaging pipel\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "99a8400e903ab330e3067629d9aa4e23ce82cf12", "messageHeadline": "[WebNN EP] Fall back resize nearest mode for WebNN CPU backend (#19039)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "52e560144978d453c73198c93043c3f1b8a30d04", "messageHeadline": "[QNN Nuget Pipeline] Build with ML ops and detect ORT version (#19024)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 4, "deletions": 9}}, {"node": {"oid": "e8ac97c8d864eb3088cf87732b5fc0a7d7df495f", "messageHeadline": "Move Windows GPU training job to A10  (#19041)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "db3c07608130e4853bb6d9db66fcf57f95a864e9", "messageHeadline": "[ROCm] do not use failed miopen fusion compile (#19012)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "4190c29d2260bb8f274049d7eab2a634fae95a21", "messageHeadline": "Add MatMulNBits accuracy_level parameter to quantization utilities. (\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 64, "deletions": 21}}, {"node": {"oid": "efdcefcf8cea1b724dd7694a8acb62fa14268e83", "messageHeadline": "[ROCm] fix security warning (#19017)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 0, "deletions": 6}}, {"node": {"oid": "447a3a7c706495fdc0f8dae8b1a130ef73af18e1", "messageHeadline": "[js/webgpu] Fix Expand/Gather when input type is bool (#18999)", "author": {"name": "Jiajie Hu", "email": "jiajie.hu@intel.com", "user": {"login": "hujiajie"}}, "additions": 48, "deletions": 7}}, {"node": {"oid": "7f0aac0d8a5cfc8fa08c3f51b9f8002ec6eed4ba", "messageHeadline": "Revert \"[WebNN EP] Rename op logicalNot to not\" (#18997)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e155c66b4a3df8a173ce3b16b31707aefec7b052", "messageHeadline": "Change all macOS python packages to use universal2 (#19013)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 94}}, {"node": {"oid": "e10a8ae31feba949b682f2451268c0dc68589ba3", "messageHeadline": "reduce max/min 20 (#17805)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 737, "deletions": 142}}, {"node": {"oid": "55a669409a93e1ed8d8e7e27b17048ce8a0cfebb", "messageHeadline": "Merge pull request #18983 from microsoft/WindowsAI", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 1772, "deletions": 362}}, {"node": {"oid": "02b1ff5fa2c41dc026022ca29c9249628f71f026", "messageHeadline": "[QNN EP] Support multithreaded inference of a single session (#18981)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 292, "deletions": 56}}, {"node": {"oid": "658e30eb33f157dc7e7cba0e6ac9bf37178722e1", "messageHeadline": "Remove DORT since it's in PyTorch main now (#18996)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 42, "deletions": 861}}, {"node": {"oid": "889b1ef2d1eca39e1216e0ca06684accf5908500", "messageHeadline": "Fix schema type constraint for custom operators (#17497)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 478, "deletions": 49}}, {"node": {"oid": "7401b6661d29094c1cb08d117fcaf8af7038b16f", "messageHeadline": "Update OperatorKernels.md", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 10, "deletions": 5}}, {"node": {"oid": "011b562b51de61c81ab8013415c7d40f677d0825", "messageHeadline": "Update c# dependencies (#18995)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "8ea3e68192967314f80ef4e7abd21a59b7cfd722", "messageHeadline": "Update ContribOperators.md", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "b18abaaa2c37e251eb639d740057aaa75821ba96", "messageHeadline": "[js/web] wait for threadpool initialization (#18952)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "867b9d8f04c6efd24b0e226f206f100f9e8bb540", "messageHeadline": "[js/webgpu] Fix f16 errors for ConvTranspose2D (#18986)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "4e2d88b75fd04f12e8e4dbf972f506c6fd0cf14d", "messageHeadline": "Remove useless `NodeProto` serializations (#18791)", "author": {"name": "Atanas Dimitrov", "email": "70822030+neNasko1@users.noreply.github.com", "user": {"login": "neNasko1"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "f4ad940ff3c69e139c3baf91517dfd458e166fb1", "messageHeadline": "Disable MatMul QDQ selector on DML EP until MatMulIntegerToFloat is r\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "d5628f52dfda37dd9c7bc3d813586c39dd13ec03", "messageHeadline": "link to docs incorrect for js/web/node (#18960)", "author": {"name": "Steven Roussey", "email": "sroussey@gmail.com", "user": {"login": "sroussey"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5fade70b5052efae1553e8e3ac0b06a527877ef0", "messageHeadline": "Update README.md (#18963)", "author": {"name": "JJ", "email": "103335846+computerscienceiscool@users.noreply.github.com", "user": {"login": "computerscienceiscool"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8e9188e265622bc811bab735c08135eec6cbd6fc", "messageHeadline": "Add SessionOptions use_deterministic_compute to the C and C++ APIs. (\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 36, "deletions": 1}}, {"node": {"oid": "70a6f816af1a482b65015616d56b73d21a43d7fe", "messageHeadline": "Port attention query fix from b2768bbf2347b4ea564f2a937f9f48987620ddf0", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "56fcea94e3c6f158f3a16dbe42dc1b16caf35565", "messageHeadline": "Enable QDQ quantization for DML EP (#18367)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 67, "deletions": 49}}, {"node": {"oid": "ee60e3af6c0e12ea3ab4a17499122c289081e92f", "messageHeadline": "Limit size of constant nodes creates by DML EP following deduplicatio\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 13, "deletions": 12}}, {"node": {"oid": "70d3f682a7a7553f337bd0422c3b8fc555af52a1", "messageHeadline": "De-duplicate 1D scale and zero point tensors to scalars in DML kernel\u2026", "author": {"name": "tbqh", "email": "111796392+tbqh@users.noreply.github.com", "user": {"login": "tbqh"}}, "additions": 153, "deletions": 9}}, {"node": {"oid": "bdaeebd6ff5d4667f288ebb9b8072f66b45de1c8", "messageHeadline": "Fix bug in DML EP ExecuteCommandList fast path and simplify design (#\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 46, "deletions": 67}}, {"node": {"oid": "b2f81c8725a0dd1ba343c91ff45aa082e9331f5d", "messageHeadline": "Hide Col2Im registration behind DML_TARGET_VERSION 6300 (#18829)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "d2f7a5b1286e34ea55dceccff8f17a45f2f799aa", "messageHeadline": "Cherry pick fix constant pow (#18785)", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "107d7492b9c0f82dc61974065c08c91550d0dea4", "messageHeadline": "[DirectML EP] Add DML EP registration for Col2Im (#17786)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 158, "deletions": 8}}, {"node": {"oid": "c1ec3c3f9369d6a01b727d4d1ef772c9fe6a5213", "messageHeadline": "User/chrila/fix dml dx12 warning (#18746)", "author": {"name": "Christian Larson", "email": "chrilaMSFT@users.noreply.github.com", "user": {"login": "chrilaMSFT"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "623d957607b571d8dadbba6d57021546c38658a4", "messageHeadline": "register resize with uint8/int8 support (#18647)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "e8209ce2b0f966739dfbb6ed4486ce78ca3f2e39", "messageHeadline": "CP 7fd1ce95a4e4f3c2b6152dfc2b1807a983ef45e5 (#18560)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 113, "deletions": 46}}, {"node": {"oid": "7f9e6c42c2e7800c8a9d38cd12e0eac881c81dd5", "messageHeadline": "readd npu enumeration (#18437) (#18518)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 39, "deletions": 8}}, {"node": {"oid": "613fdce12e528d6921d0c2c7674e0fbbe2c18e6a", "messageHeadline": "Create ring buffer for re-used command lists (#18368)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 68, "deletions": 50}}, {"node": {"oid": "5c283340c384d5f9e402ba762c3c33af14e6763e", "messageHeadline": "Filter activation fusions on MCDM (#18371)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 50, "deletions": 20}}, {"node": {"oid": "a1000a0a3c2552b78045bb5452aea09ae3490e6a", "messageHeadline": "Enable GEMM activation fusions on MCDM (#18372)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "531e875fb550f7a866cd13639fb2e332440e6b96", "messageHeadline": "Avoid command list reset in common case of re-used command list execu\u2026", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 18, "deletions": 0}}, {"node": {"oid": "d5f3aae3fd30a79d9bbeed26a70907618ac84835", "messageHeadline": "Utilize DML constant input graph node (#18267)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 229, "deletions": 38}}, {"node": {"oid": "dcfff10f57fbbdf81ea9ebcae008be712be42700", "messageHeadline": "Enable QLinearAveragePooling DML EP (#17384) (#18240)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 339, "deletions": 6}}, {"node": {"oid": "cb7f28a16ab78601363c2e694679d2dada149dd1", "messageHeadline": "Register Resize for INT8 and UINT8 (#18252)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9ff5e3b7b0f5e9683422fa3609175fe4f82ccb77", "messageHeadline": "Add QLinearConcat for DML EP (#16971) (#18268)", "author": {"name": "raoanag", "email": "127366241+raoanag@users.noreply.github.com", "user": {"login": "raoanag"}}, "additions": 290, "deletions": 13}}, {"node": {"oid": "9bbe425d7f805d4dfaad2e69c3edca40063fe673", "messageHeadline": "Register LPpool18 and AvgPool 19 (#16880)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 153, "deletions": 29}}, {"node": {"oid": "c3d96a7b35c975a4eb4ad5a4c94349797defbc78", "messageHeadline": "Update DML version to 1.13.0 (#18978)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "3b8b9147fa4f8f6348e171a257bbc325744301df", "messageHeadline": "[js/webgpu] Mitigate floating point accuracy issue in Resize (#18956)", "author": {"name": "Jiajie Hu", "email": "jiajie.hu@intel.com", "user": {"login": "hujiajie"}}, "additions": 45, "deletions": 38}}, {"node": {"oid": "c5f3952b68872ac80d8b9a45a57d46262a455ffd", "messageHeadline": "[js/webgpu] Introduce trace support (#18928)", "author": {"name": "Yang Gu", "email": "yang.gu@intel.com", "user": {"login": "gyagp"}}, "additions": 75, "deletions": 2}}, {"node": {"oid": "7a454acd6197f4ba1ffca13ec9948915ce82d20e", "messageHeadline": "[ROCm] Update CI/Packaging pipeline to ROCm6.0 (#18985)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 35, "deletions": 35}}, {"node": {"oid": "c97e3f48216d66dfbc6aa951ddcb7f32e313d314", "messageHeadline": "[Fix] exception in Fuzz Test pipeline (#18984)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "1c2dca95d813e3bf7a2b59a70fcedae9c84bed7d", "messageHeadline": "pass rotary embedding to attention op (#18846)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 45, "deletions": 25}}, {"node": {"oid": "df740d7d152dfd04d29f4310a02d41b68f010b37", "messageHeadline": "Throw if unique_ptr or array allocation fails due to SafeInt overflow\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 75, "deletions": 32}}, {"node": {"oid": "3993d4304888d5622d929b8912b160ffd00b3f94", "messageHeadline": "[EP Perf] Fix missing Azure cli & use onnx zoo model inside image  (#\u2026", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "81cbdb10a98fa0456e867f797a2f554b2021d0af", "messageHeadline": "Bump actions/upload-artifact from 3 to 4 (#18920)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "780fc3611b6121222c53133574d194545bed3dbf", "messageHeadline": "[JS/Web] Sajandhy/webgpu resize scales rank check (#18954)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 54, "deletions": 4}}, {"node": {"oid": "96d1f3203a0453c295754302f9b60aff17d498a8", "messageHeadline": "[WebNN EP] Decompose Concat with input number > 4 for CPU backend (#1\u2026", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 31, "deletions": 4}}, {"node": {"oid": "a3626b67b31909c4109aa38bb079f7b5df047b72", "messageHeadline": "[WebNN EP] Rename op logicalNot to not (#18936)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "44584c3ebe7058ce78eb6c1ceebec6697885302b", "messageHeadline": "[js/webgpu] only declare shape and strides in shader when necessary (\u2026", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 13, "deletions": 4}}, {"node": {"oid": "c613cc58a9e5001434d2f562c69fc432e6196ebd", "messageHeadline": "[js/webgpu] Fix shader compilation errors in Resize (#18947)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "998517b209b377b30f923f0a18a971b6f25f760f", "messageHeadline": "Minor fixes (#18949)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "3bbe4fe2ff27da46944f526481d64b9c964a3066", "messageHeadline": "[JS/WebGPU] Add trilinear interpolation to Resize; activation_params \u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 254, "deletions": 51}}, {"node": {"oid": "31d4a21c4bff39946f3884ab3fec90b1d4d5224e", "messageHeadline": "[js/webgpu] fix heap access > 2GB (#18914)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a90526667d720d46119fbf0faa6d190c633a1c50", "messageHeadline": "Bump actions/setup-dotnet from 3 to 4 (#18919)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e08bdc109d6a6cda6f8628be5cf8f68367ac1970", "messageHeadline": "Bump github/codeql-action from 2 to 3 (#18922)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "437d339d66d376a5eee05ee09e84f061afab2d2a", "messageHeadline": "Bump actions/setup-python from 4 to 5 (#18921)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "0bc71b0c9b817b0e54d9b6f1a7f92eaa1b77ea66", "messageHeadline": "[js/webgpu] Refactor attributes of pool (#18728)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 33, "deletions": 28}}, {"node": {"oid": "dbb8680bdc8badd0616cc7fac9d9f82231188be3", "messageHeadline": "Delay load dxcore.dll in addition to ext-ms-win-dxcore-l1-1-0.dll (#1\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0c5826f013023c477e7342c2de0cf61c2c9b7709", "messageHeadline": "attn_mask fix, set as torch.long (#18931)", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "37f743680a0690270292d917b23509993802d7c1", "messageHeadline": "Fix build when flash attention and memory efficient attention are dis\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 115, "deletions": 10}}, {"node": {"oid": "9dd9461e655088240cc09d3541cbf7f5c6bb91f9", "messageHeadline": "Fix crash in PadFusion (#18557)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "3f03c12986b92ab3916d72f9ecd74c836f0d2a63", "messageHeadline": "Split Onnxruntime Nuget GPU package (#18819)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 295, "deletions": 136}}, {"node": {"oid": "985acda28ca1f43fd34cb65039c8ad8cdab9ec0f", "messageHeadline": "optimize int4 gemv kernel with cuda (#18818)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 217, "deletions": 76}}, {"node": {"oid": "3d8f229d3938dbd467b713daf5d8a4d855ca53bc", "messageHeadline": "Add ARM64EC build jobs (#18870)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 46, "deletions": 1}}, {"node": {"oid": "5b93c465a87bf65bf299effe685d7c39a52d2378", "messageHeadline": "Delete .github/workflows/generated_fake_win_gpu_ci.yml (#18074)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 36}}, {"node": {"oid": "0af946f35add15dc3062f139d8e557dcacf64f90", "messageHeadline": "[EP Perf] Fix ORT-CUDAFp16 tests (#18908)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "4c3705cbea028f5abd061b46833884bfa4403b9d", "messageHeadline": "[WebNN EP] Change some support status for XNNPack backend (#18858)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "1b64d30963eeb7419b58bfe1ef2e6d27256d7043", "messageHeadline": "[WebNN EP] Infer the layout via ONNX domain for Resize (#18871)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 18, "deletions": 11}}, {"node": {"oid": "8507c06f8eb428f1b8b93b044152f569bb387ed1", "messageHeadline": "Bump conda-incubator/setup-miniconda from 2 to 3 (#18685)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "914bc409b0e1258db3980ff881a9d4beb6403fbd", "messageHeadline": "Bump transformers from 4.30.0 to 4.36.0 in /tools/ci_build (#18895)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f3c62bfad9b7bd4f0a90633df26e4774cfb699ff", "messageHeadline": "Bump actions/setup-node from 3 to 4 (#18148)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f74389c976a8ee445f9359a6d302756b846ca739", "messageHeadline": "Bump github/issue-labeler from 3.2 to 3.3 (#18408)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "54e471a054fef814cdd675f99af8fde7fa6bb728", "messageHeadline": "[EP Perf] Display percentage of cuda/trt ops in cuda/trt ep on EP Per\u2026", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 280, "deletions": 236}}, {"node": {"oid": "ce70a30b94012eb1652696ff7cc7a8f62cd20c60", "messageHeadline": "Bump transformers from 4.35.2 to 4.36.0 in /onnxruntime/python/tools/\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "379c7c43eb96595028b63df15a8236c16773a80d", "messageHeadline": "Bump actions/setup-java from 3 to 4 (#18686)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1c6cb5dfebe38b09627430f0add54416653a2ceb", "messageHeadline": "Remove usage of TRT deprecated APIs (#18879)", "author": {"name": "Kevin Chen", "email": "45886021+kevinch-nv@users.noreply.github.com", "user": {"login": "kevinch-nv"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "2d6e2e243d1a1ab0486f4f191b61ac979c5b978e", "messageHeadline": "update sdxl demo (#18889)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1305, "deletions": 2524}}, {"node": {"oid": "9a61388f0a3644fc2b572e2164446550ce9b1aa6", "messageHeadline": "[js/web] revise backend registration (#18715)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 393, "deletions": 390}}, {"node": {"oid": "c0142c9108120e40141d112009e6ab7a8810a63f", "messageHeadline": "[EP Perf] Fix model zoo url (#18808)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8931854528b1b2a3f320d012c78d37186fbbdab8", "messageHeadline": "Move some QNN EP provider options to session options (#18877)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 171, "deletions": 104}}, {"node": {"oid": "02eb17655d8d76d078bf61af438f4baa5e62b375", "messageHeadline": "Fix a bug in 4bits quantizer script (#18878)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "666fcbde4d8375ff6434053a9ae6ca34719c6beb", "messageHeadline": "Add LeakyRelu to list of NNAPI operators (#18880)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "535a2403dd74ef8d694706c5962acae179fcaca4", "messageHeadline": "Update Nuget publishing jobs (#18851)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 96, "deletions": 49}}, {"node": {"oid": "ffa660268606b7f422acb7fee109d91fa90c7191", "messageHeadline": "[js/node] support manually dispose session (#18655)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 30, "deletions": 2}}, {"node": {"oid": "98510fb8fb2d761aad73ed8d32d4a6922546bdf0", "messageHeadline": "[JS/WebGPU] fix an error in Clip (#18799)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 10, "deletions": 11}}, {"node": {"oid": "32fcf73740aba943d4dc3a436838428c6b20163d", "messageHeadline": "Implement dft(20) (#17821)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 101, "deletions": 37}}, {"node": {"oid": "5f00bc99311a5d4eb7b0269c1f1cf1a7db1a1f9a", "messageHeadline": "Integrate high-performance x64 gemm library to MLAS (#17669)", "author": {"name": "luoyu-intel", "email": "yu.luo@intel.com", "user": {"login": "luoyu-intel"}}, "additions": 24902, "deletions": 10}}, {"node": {"oid": "4dff154f51d8e1fa4db63729a5c0796494886d6c", "messageHeadline": "Fix nightly pipeline failure (#18867)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "6d7519ede8298a422e84e70bfdf01cc46fbf76c3", "messageHeadline": " Adding new pipeline for python cuda testing (#18718)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 156, "deletions": 1}}, {"node": {"oid": "63b47ceaf892da6d960df73a3ab0007be9a1e8ef", "messageHeadline": "[REACT NATIVE] Bugfix -> casing Podfile (#18861)", "author": {"name": "Frank", "email": "frankbaele@users.noreply.github.com", "user": {"login": "frankbaele"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3ff4a4c393dad8c67fa6019c87b844e4981b0a11", "messageHeadline": "Bump actions/stale from 8.0.0 to 9.0.0 (#18774)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ea6186efa8e0fd9b1b62a8c392508af088e9df8e", "messageHeadline": "Update stale.yml to correct close-issue-message (#18849)", "author": {"name": "sophies927", "email": "107952697+sophies927@users.noreply.github.com", "user": {"login": "sophies927"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9426bd50cb52cd0715e5f917cc70bff3190ef4c1", "messageHeadline": "[TensorRT EP] Update deprecated TRT api (#18834)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 18, "deletions": 7}}, {"node": {"oid": "ad476d5a1fb63a4cad74899873ccbf61e9487a23", "messageHeadline": "Change Nuget packaging pipeline's build TRT job to download CUDA SDK \u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "50cbcf95877b60795f32c4538611f9a119bb0291", "messageHeadline": "Build function bodies according to the imported global opset. (#18833)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 77, "deletions": 12}}, {"node": {"oid": "2952cf82a52ade99fee9ee9dcfd3570dd4e51863", "messageHeadline": "Access map by iterator to silence sanity check. (#18835)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "8f7b89bd5bbfce6983dbd1935e7073bad7701921", "messageHeadline": "[js/webgpu] Optimize NCHW layout for InstanceNormalization (#18123)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 23, "deletions": 19}}, {"node": {"oid": "4bbed4c71a38f9a7db8e5f0ce4385f30fa4d2338", "messageHeadline": "[js/webgpu] Fix f16 errors in unary (#18839)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 16, "deletions": 12}}, {"node": {"oid": "f52668cc68efe80197227da192d9b970fa739132", "messageHeadline": "Disable mlas unit test in ARM64EC build (#18747)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 47, "deletions": 48}}, {"node": {"oid": "89168b830d663647c00fd74536aee52f0671f884", "messageHeadline": "Fix CI error:  The workflow is not valid. .github/workflows/rust-ci.y\u2026", "author": {"name": "wirthual", "email": "wirthra@gmail.com", "user": {"login": "wirthual"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "81ad1e6ac3149b928ccdaed9f76195a303613804", "messageHeadline": "[js/webgpu] Fix typo of outputShapes in profiling message (#18837)", "author": {"name": "Yang Gu", "email": "yang.gu@intel.com", "user": {"login": "gyagp"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d111eed726f6009bd9c4bf3355194a3b85aabb9f", "messageHeadline": "[WebNN EP] Change axis to axes for argMax/argMin (#18838)", "author": {"name": "Peishen Yan", "email": "peishen.yan@intel.com", "user": {"login": "peishenyan"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "d795fc636ce92c29d95d85cf0faf506baeadd46b", "messageHeadline": "FIX: Our cmake script didn't check googletest's hash (#18826)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "fc9ecb59dbf6ac647bb1a70727a45e9267fefa90", "messageHeadline": "Add Windows ARM build jobs to post merge pipeline (#18832)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 144, "deletions": 6}}, {"node": {"oid": "5eda79bdd3f138d599d5d0dda75b76096ea62a93", "messageHeadline": "Improve perf for stage3 training (#18099)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1520, "deletions": 1183}}, {"node": {"oid": "cbad4fe49bfada781059659f555fcde49fbae37f", "messageHeadline": "Update absl and googletest (#18827)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "b42d4b8ea650c7b384bfbac1c7edc292c60747a6", "messageHeadline": "[VitisAI] 1. api compatbile 2. dynamic load onnx (#18470)", "author": {"name": "Yueqing Zhang", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 199, "deletions": 293}}, {"node": {"oid": "6d5ee4d69bd7aac085bd8dca5a391227e628948d", "messageHeadline": "[WebNN EP] Use explicit padding (#18688)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 111, "deletions": 121}}, {"node": {"oid": "1db1c750488cd6602ea2fa741678b5bd1b16da5f", "messageHeadline": "[WebNN EP] WebNN only supports 4-D input and weight for Conv/ConvTran\u2026", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 30, "deletions": 13}}, {"node": {"oid": "b129f425fcf450ce382f7caba2b564e7c3d47f3f", "messageHeadline": "Fix test model URL issue (#18823)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "afe5cdc9387ab58c383a62a2d3b3f4a74dac532d", "messageHeadline": "[TensorRT EP] Switch to enqueueV3 with support DDS output (copy versi\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 619, "deletions": 313}}, {"node": {"oid": "7386e211218d9c2a1d852659cf22de908d7ad898", "messageHeadline": "Replace some ORT_ENFORCE with ORT_THROW_IF_ERROR (#18812)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 34, "deletions": 35}}, {"node": {"oid": "95193cb440128570891df3d281be6415e9cf1dd8", "messageHeadline": "Set NDK version in Linux CPU Minimal Build E2E CI Pipeline (#18810)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "7dade5d05b67f4da8cc9ab949d576159682aff20", "messageHeadline": "Readd basetargets in Microsoft.ML.OnnxRuntime.csproj (#18789)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7047d13c68652044cb24aebaa71ab362f8b0a7b4", "messageHeadline": "Update windowsai-steps.yml: enable \"/profile\" linker flag (#18022)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0723dcb8b591a559db60885ff2cad610fd989ad4", "messageHeadline": "OpenVINO Execution Provider with 2023.2 support (#18596)", "author": {"name": "Suryaprakash Shanmugam", "email": "suryaprakash.shanmugam@intel.com", "user": {"login": "sspintel"}}, "additions": 141, "deletions": 94}}, {"node": {"oid": "f3fa0456815c78474be36bb2e9a7e18f6b703aa8", "messageHeadline": "Enable MacOS build in ORT Objc Pod (#18786)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "487abcd25ec2bcb2255a361e4b061f020a90c043", "messageHeadline": "Update gradient ops tests (#18783)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "17eaf9b053238b3efec303e9c94008201ca42462", "messageHeadline": "Fix a build warning in SparseTensor code for 32-bit build configs (#1\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 50, "deletions": 30}}, {"node": {"oid": "44054e7508b4a37748213585eb644faef013ddf1", "messageHeadline": "Move NuGet nightly package publishing job to a separated pipeline (#1\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 35, "deletions": 44}}, {"node": {"oid": "b30e721dc874c8e32cb3ce6fd0b00b63ac3716ff", "messageHeadline": "[js/webgpu] Provide a naive vectorized matmul algorithm (#18758)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 164, "deletions": 10}}, {"node": {"oid": "1ad6eb135959028bcc0346206c6a8b5cf17d16ee", "messageHeadline": "Add DynamicQuantizeLinear as supported OP (#18798)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "dbe886abb3b3615a478a37a1806f9107018eb49b", "messageHeadline": "Disable test_bert_result_with_layerwise_recompute (#18800)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "3940ef20beca9aa47ed0e36b200f121673f33482", "messageHeadline": "[ROCm] Refactor to hide ck layout (Row/Col) from ORT interface (#18777)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 262, "deletions": 255}}, {"node": {"oid": "0ca84549abac23aa9c9347df1a3ab68cee9c02b1", "messageHeadline": "[JS/Web] Added uniforms to Reduce, Resize and Split Ops. (#18727)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 219, "deletions": 204}}, {"node": {"oid": "81796a30810ca9038474260742e542fffa11fc71", "messageHeadline": "[QNN EP Quantization] Add fusion preprocessing to QNN quantization (#\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 953, "deletions": 5}}, {"node": {"oid": "65300610e2df35a2371f6cb5292a8f030fc409ea", "messageHeadline": "[PowerPC] Type casting the output operand of vec_xst. (#18057)", "author": {"name": "BODAPATIMAHESH", "email": "148746454+BODAPATIMAHESH@users.noreply.github.com", "user": {"login": "BODAPATIMAHESH"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d673e39ad89a709d5896510bcd496927567b4b79", "messageHeadline": "[JS/WebGPU] Added uniforms to Tile and Where Ops (#18768)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 47, "deletions": 39}}, {"node": {"oid": "b4be9e1bbb20e1e03528f73df71e9f141ae04fcf", "messageHeadline": "[js/webgpu] Fix shader compilation errors in cumsum (#18779)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 37, "deletions": 1}}, {"node": {"oid": "a85ef652ed0c0626fe04d1a7da3574f7f466c22e", "messageHeadline": "Log out ORT session options (#16259)", "author": {"name": "ivberg", "email": "ivberg@microsoft.com", "user": {"login": "ivberg"}}, "additions": 169, "deletions": 1}}, {"node": {"oid": "eb030329257e1859eaa0e27c61b7c68517c960d2", "messageHeadline": "[js/web/training] lazyResetGrad implementation (#18711)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 27, "deletions": 1}}, {"node": {"oid": "ccf3b2054b47c3a48001bd9305957d430ac02f0e", "messageHeadline": "Allow layer-wise recompute (#18566)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 931, "deletions": 231}}, {"node": {"oid": "68c832d53bfc1965730103fdc94019e8155ea348", "messageHeadline": "Fix buffer overrun in 4b dequant cuda (#18780)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "ce1fed6ddf649b0e2d0428525449f9152b132d59", "messageHeadline": "Adding a new pipeline for publishing to Python Cuda 12 packages. (#18\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 75, "deletions": 0}}, {"node": {"oid": "bfa5eb4591fed374c07a8e9e8eda2ec4c682b3e2", "messageHeadline": "Adding a new pipeline for pubilshing cuda 12 nuget packages (#18713)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 83, "deletions": 0}}, {"node": {"oid": "16df8377d39308237ec2909f178a137ddd9a0a80", "messageHeadline": "Update transformers package to fix the security issue (#18730)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 29, "deletions": 25}}, {"node": {"oid": "8d641229e6dbd6364a610923c31fc51448e2601a", "messageHeadline": "Fix GQA shape inference (#18723)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "de32baeeeff6ec8dc4f0ac8edbf4a46436eb7991", "messageHeadline": "[ROCm] Add GemmFloat8 (#18488)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 1648, "deletions": 36}}, {"node": {"oid": "d41dd772416f55844d2051a4050a0df439826797", "messageHeadline": "Extend API page on the python documentation (#18762)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 74, "deletions": 0}}, {"node": {"oid": "2f93d97fd02e9d096179fb6c4215b2614c3ce42a", "messageHeadline": "Add cuda visible devices for Mistral benchmark (#18764)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "c7799d70585ec1455e013c61b280b044a7a73b15", "messageHeadline": "Build fixes for Windows ARM32 desktop build (#18752)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 1}}, {"node": {"oid": "44b58437402b207c8216f3be8c75accb7409be1c", "messageHeadline": "Fix gemm_float8 build failure on CUDA 11.3-11.7 (#18760)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "e8f33b54bab5129b0dea177669bbd1c1d0894dd8", "messageHeadline": "[WebNN EP] Don't covert all inputs except the 0th input for Resize (#\u2026", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 4, "deletions": 11}}, {"node": {"oid": "7ed48a299a5d81a3baef39bfe3327fbccb85eff1", "messageHeadline": "Objective-C API updates (#18738)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 45, "deletions": 4}}, {"node": {"oid": "bf33919afba1fe55258f644f3136fb073a85b2c2", "messageHeadline": "Update absl and gtest to fix an ARM64EC build error (#18735)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 31}}, {"node": {"oid": "305db31301e97e940f42f6c9642f6d1f0aebc9bc", "messageHeadline": "fix build aar error in Zip-Nuget-Java-Nodejs Packaging pipeline (#18745)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "efbef5f6115c0156f3ea3cc348bd2e57f293d241", "messageHeadline": "[js/webgpu] allow to specify callback for profiling data (#18732)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 71, "deletions": 22}}, {"node": {"oid": "4abec9749e0cd3bcd22ed3025d8505f91e80f562", "messageHeadline": "[mlas] add loongarch lsx and lasx optimize code (#17937)", "author": {"name": "junchao-loongson", "email": "68935141+junchao-loongson@users.noreply.github.com", "user": {"login": "junchao-loongson"}}, "additions": 7696, "deletions": 34}}, {"node": {"oid": "a045be335b06f7b26b24b1b51e43e52a83ffa2bc", "messageHeadline": "use EO pool for windows web_cpu stage (#18737)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e469de65f5eab2089b6273e7acc5e37bd645bd89", "messageHeadline": "Re-enable Sign op int64 test for QNN CPU test (#18734)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 1, "deletions": 10}}, {"node": {"oid": "3d8af6eb65c0507ec491307917aaa37665c3cd24", "messageHeadline": "[WebNN EP] Skip split initializer (#18729)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "49470f06e88ff99837e7ab0ae6062c32a782e068", "messageHeadline": "Add benchmark script for control net (#18717)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 302, "deletions": 6}}, {"node": {"oid": "e603e78627ac2765301e0f8e9a5f76f8fb2fe9ec", "messageHeadline": "Enforce If condition size == 1 (#18733)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "9479ba525b55dbbb4bf2bf4e18ce74c70ecf3171", "messageHeadline": "Build onnxruntime.dll as arm64x (#18633)", "author": {"name": "moyo1997", "email": "54333118+moyo1997@users.noreply.github.com", "user": {"login": "moyo1997"}}, "additions": 61, "deletions": 0}}, {"node": {"oid": "7762f3f7c550d05c7a053843b988951219de7b44", "messageHeadline": "[NNAPI EP] Add NNAPI Split (#18702)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 167, "deletions": 12}}, {"node": {"oid": "c4b8120c5b77bb1a7fd708b3a1804fb5ad49446e", "messageHeadline": "Rename op elementwiseIf to where (#18657)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9768a727e1006b84673f818924fee20b5c4288e1", "messageHeadline": "[QNN EP] Fix a bug that can't create context binary if the model has \u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 89, "deletions": 15}}, {"node": {"oid": "559bd52252f2db17e849c9101da4a22ad6e69f8b", "messageHeadline": "[QNN EP] Update QNN SDK to version 2.17.0 (#18684)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 521, "deletions": 230}}, {"node": {"oid": "c012e41f9385303f486b644cd679fdb2784fe854", "messageHeadline": "MoE with Expert Slicing  (#18565)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 884, "deletions": 144}}, {"node": {"oid": "871c52977aa4297d783fd4d830eaa10c71cb2be6", "messageHeadline": "Mistral Optimization & Benchmarking Support (#18225)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 111, "deletions": 7}}, {"node": {"oid": "c9e558cd36bf074b04d12a1e9c2d5498f3e9fb6f", "messageHeadline": "Adding common python test requirements.txt (#18698)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "4bfa84487cc6fe992b18d69ccd5f0d54392b64f5", "messageHeadline": "Skip module clone for preparing large model export (#18663)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 76, "deletions": 6}}, {"node": {"oid": "9aa7284351ae7191fad8def3951a634ce61d0082", "messageHeadline": "fix lint error (#18708)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "07aabcc314607fa35580956ea45c0bcd1707e394", "messageHeadline": "Set cuda device before create cuda stream for IOBinding case (#18583)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 86, "deletions": 35}}, {"node": {"oid": "70816001ccae305de24e27ab2219a8a17e1ca036", "messageHeadline": "[JS/Web] AddedUniforms in GatherElements. (#18670)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 26, "deletions": 32}}, {"node": {"oid": "f949e0580b477727e1444f5a9a05bec7929ab0d7", "messageHeadline": "[js/webgpu] Support uniforms for pool (#18656)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 147, "deletions": 70}}, {"node": {"oid": "10c547516d0e65583542b356c08c349c25dc5e6d", "messageHeadline": "[JS/Web] Added CumSum operator to JSEP (#18637)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 1493, "deletions": 6}}, {"node": {"oid": "c14fae9461a18184f5e6b8d559914ff4041b947e", "messageHeadline": "add SAVE_TEST_GRAPH macro (#18696)", "author": {"name": "rui-ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 11, "deletions": 2}}, {"node": {"oid": "2b3050bb0c89537d67e213f657ec56a7ec21d47e", "messageHeadline": "Zhijxu/fix toposort (#18705)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 10, "deletions": 3}}, {"node": {"oid": "e066fca7770987c9c2c91babca9d74e95291e39f", "messageHeadline": "[Quantization] Tensor quant overrides and QNN EP quantization configu\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 825, "deletions": 56}}, {"node": {"oid": "01b5c789177c2b062d4c4f9b6abdce12be9b3b64", "messageHeadline": "Add SD-Turbo and refine diffusion demo (#18694)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 207, "deletions": 191}}, {"node": {"oid": "d514a960eefc19fb69d54497b6b582cfdf6e85f1", "messageHeadline": "Remove \"Python Checks\" pipeline status from readme as that pipeline n\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c02a3861451a29d7a517dd4aaa82c239d2f34d2d", "messageHeadline": "[js/web/training] Implemented runEvalStep & runOptimizerStep (#18259)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 242, "deletions": 61}}, {"node": {"oid": "5353adcde37a118bdd25882482fd584c5ed3f343", "messageHeadline": "[js/webgpu] Use the naive convTranspose when in/out channels are both\u2026", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "a5b2291e0fe7c7d42f30154ccb20d6cde1380c3c", "messageHeadline": "[Transformer Optimization]Return model directly for unknown model typ\u2026", "author": {"name": "trajep", "email": "jiapli@microsoft.com", "user": {"login": "trajepl"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "2f8b86b93906d0dd0549aca22798c660aa10db91", "messageHeadline": "Fix typo in the TensorShape (#17813)", "author": {"name": "Deoksang Kim", "email": "deoksang.kim@sapeon.com", "user": {"login": "deoksangkim"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "92ee664f64e96a8cc7308302a3e4f67f95254d1f", "messageHeadline": "[js/webgpu] Fix shader errors in indicesGet/Set when rank > 4 (#18661)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 22, "deletions": 18}}, {"node": {"oid": "eaaf27015e8d99c5a072caa40e0f4627f14a93e3", "messageHeadline": "Remove EnvSetupScript parameter from win-ci.yml (#18662)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 26}}, {"node": {"oid": "9c45fe4957ff3d027b5024abb170947db2cb0408", "messageHeadline": "Fix macos xcframework test stage codesign info (#18649)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 7, "deletions": 33}}, {"node": {"oid": "a3538056314c10c1c4d5b769e86426434d486322", "messageHeadline": "Fix Windows TVM CI workflow (#18667)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 5}}, {"node": {"oid": "b22f49ff35b3c7b3ae339128e21898810e4c2919", "messageHeadline": "Fix unit tests failures in build with contrib ops disabled (#18659)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 60, "deletions": 49}}, {"node": {"oid": "fcea2cb7f184d608efa1e5c72f9e25072e82009d", "messageHeadline": "[Dort] Run type promotion pass to resolve dtype discrepancy (#18516)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "05a9c957647b3cae0d2ad305950c14bf5f305bc8", "messageHeadline": "[DNNL] add Arm Compute Library (ACL) backend for dnnl execution provi\u2026", "author": {"name": "snadampal", "email": "87143774+snadampal@users.noreply.github.com", "user": {"login": "snadampal"}}, "additions": 22, "deletions": 1}}, {"node": {"oid": "d69842226b47e5336568103541b071447caeb9bf", "messageHeadline": "Update the template files to correct stage to fix the python cuda 12 \u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "182c525416eb5cbace8df52b6809a77ffc91545d", "messageHeadline": "Support MatMulBnb4 in PaddingElimination (#18646)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 25, "deletions": 10}}, {"node": {"oid": "ccfea559428b1374d0109bfaacc273ce11f4ef3c", "messageHeadline": "[QNN EP] Enable QNN HTP VTCM size setting (#18653)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 76, "deletions": 55}}, {"node": {"oid": "9c9e6adeb2f31c73cebd7e92622c86f084858f68", "messageHeadline": "Add SDXL Turbo to demo (#18627)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 402, "deletions": 133}}, {"node": {"oid": "c7732a78d7e815de489fed22cfee610a445b9ca2", "messageHeadline": "[WebNN EP] Fixed bug in op checking (#18638)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "73d9b035090a2bd4e56252dee10174d3f01e5f6f", "messageHeadline": "[js/webgpu] Add multidimensional(>4) uniform support (#18546)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 65, "deletions": 95}}, {"node": {"oid": "73a2eb82eb9364b4dea8df2cd6a46affd008b15c", "messageHeadline": "Fixed bug in Flatten's axis (#18645)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "6781b6cf3d4708e32e6bd546afa5b2b785290270", "messageHeadline": "[js/webgpu] add bool type for Expand/Gather (#18615)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 235, "deletions": 68}}, {"node": {"oid": "efee9abdb72f73163943df80f0e6db1f5c23c42c", "messageHeadline": "Reduce downloads in Nuget-Java pipeline to reduce connection exceptio\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 178, "deletions": 166}}, {"node": {"oid": "4025bd8ebdda49331af45c7632cb5975fedf69c2", "messageHeadline": "[WebNN EP] Fix bug of padding in Op ConvTranspose (#18577)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "b1e749e3beb8fe543500f7ba51ddc9754639525d", "messageHeadline": "[js/webgpu] Add program name into webgpuProfiling info (#18640)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c5ea1547c6d1070e6b6296fbf8e6d681107b8c7f", "messageHeadline": "Eliminate intermediate string conversion buffer. (#18608)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "e7f64f4510483bf0a94ce46478f02ead8d70e0d2", "messageHeadline": "[js/web] fix ESLint by excluding generated .js from tsconfig.json (#1\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "23a91c8ba889d77589d6acf44fa9e9bce5fbb701", "messageHeadline": "Fix warning C4003 in ORT python binding code (#18612)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 22}}, {"node": {"oid": "1b5675ff0fc7b2d9894ef06a7727efe0aad7cbd2", "messageHeadline": "Update post-merge-jobs.yml: increase timeout value for the Ios job (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "148495ebc55827c8c521ea41493052ddbc428ab2", "messageHeadline": "[ORTModule] Use Default Topo-order for GraphViewer (#18410)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 39, "deletions": 3}}, {"node": {"oid": "e1d1033131114dc2634e664d009e061d900a9554", "messageHeadline": "[ORTModule] Remove Unused Arguments from Generated Triton Code (#18636)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 53, "deletions": 14}}, {"node": {"oid": "5c67a00d8e9ba3604593b6fe25a1e3da0c8ef65b", "messageHeadline": "Revert \"remove full protobuf requirement for tensorrt ep\" (#18626)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "c20488ced70488c9e95b6c11fdea309efe2fdc99", "messageHeadline": "skip_infer for SkipGroupNorm in SymbolicShapeInference (#18630)", "author": {"name": "Jambay Kinley", "email": "jambaykinley@microsoft.com", "user": {"login": "jambayk"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "227dcb3a88eb8c36bfc5c0341156ce96291597ac", "messageHeadline": "[js/webgpu] Log the key and program info for artifact (#18365)", "author": {"name": "Yang Gu", "email": "yang.gu@intel.com", "user": {"login": "gyagp"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "7335760424b052ff041285571cf52b77f9ebb009", "messageHeadline": "[JS/Web] Add uniforms to Einsum (#18531)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 453, "deletions": 97}}, {"node": {"oid": "483c490ec4db2d2b5001e42f5c842abfc9e379af", "messageHeadline": "Refine error checks in onnxruntime/core/providers/coreml/model/model.\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "d2dfbf41795e72911643e2ffcadac069b72580bd", "messageHeadline": "Add float16 type support to SplitToSequence and make code type indepe\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 114, "deletions": 83}}, {"node": {"oid": "68209307daadfe21a74a36d44c4c170b91141772", "messageHeadline": "Replace all Azure-Pipelines-EO-Windows2022-aiinfrat to Onnxruntime-Wi\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 9, "deletions": 11}}, {"node": {"oid": "38b640c797613e2396f2975ccd4d8ff0e95a5baa", "messageHeadline": "[WebNN EP] Re-implement Unsqueeze, Squeeze, Flatten with WebNN's resh\u2026", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 49, "deletions": 20}}, {"node": {"oid": "14a343441dcd530bec24e18e34c3c068993eb06c", "messageHeadline": "Fix Objective-C static analysis build (#18606)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 51, "deletions": 18}}, {"node": {"oid": "e833d22f143f86529f4863b5da6cac4eb4a78bbb", "messageHeadline": "Change QNN EP Profiling logs to output to CSV (#18201)", "author": {"name": "ivberg", "email": "ivberg@microsoft.com", "user": {"login": "ivberg"}}, "additions": 227, "deletions": 17}}, {"node": {"oid": "f13380f3d8d25df797be60b4899b43504a5576b5", "messageHeadline": "Support LoRA and Control Net in Stable Diffusion demo (#18593)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1044, "deletions": 212}}, {"node": {"oid": "50e6235af111e5113860dfd7a0ece55dc00316a0", "messageHeadline": "[js/web] allow ShaderHelper to use internal (non-I/O) variables (#18525)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 83, "deletions": 51}}, {"node": {"oid": "a49f31b6705bdd8a9b9cd7b7b4a9bbc0ebba07a2", "messageHeadline": "Remove drop-nuget artifact from all pipelines (#18592)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 2, "deletions": 17}}, {"node": {"oid": "e24733cfe9b3e0d40419942f2d6337925c351606", "messageHeadline": "fix the Olive CI pipeline failure on Windows (#18464)", "author": {"name": "Mike Guo", "email": "guotuofeng@gmail.com", "user": {"login": "guotuofeng"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "288b80d363bc120c8d3c0ca3c2fe4252e16f4c56", "messageHeadline": "Add MacOS build to ORT C Pod (#18550)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 1512, "deletions": 197}}, {"node": {"oid": "05046e5452f7a1f47bb1f4c01ddfa86eb6fac77f", "messageHeadline": "Adding unit test for sm80 prepack (#18514)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1337, "deletions": 5}}, {"node": {"oid": "8d5ecc4dae0686d032a81c3633fdaf213572a722", "messageHeadline": "[Quantization] Fix scale/zero-point for 16-bit QDQ Softmax (#18589)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 93, "deletions": 34}}, {"node": {"oid": "0b7048e7d621b271b0ab4748e566f57d11b49be5", "messageHeadline": "Update winml to use #cores - #soc cores by Default as the number of i\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 117, "deletions": 10}}, {"node": {"oid": "a6d872640764ea50ec460f7a717e5b369921f8b4", "messageHeadline": "Update ADO windows image to custom image (#18598)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 72, "deletions": 54}}, {"node": {"oid": "3ea27c29253aad7c02015e2af6d37dedafe2c9c3", "messageHeadline": "Create a new Nuget Package pipeline for CUDA 12 (#18135)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 889, "deletions": 53}}, {"node": {"oid": "94a6020a7f59f22101653988a36bca02593eb816", "messageHeadline": "Improve parallelization of TfIdfVectorizer, Reduce memory consumption\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 71, "deletions": 90}}, {"node": {"oid": "3f42fbad2e42cf03c01eb0428b06e24f4ad2d427", "messageHeadline": "deleted the unused random_device variables because they caused a warn\u2026", "author": {"name": "Ran Gal", "email": "79867742+galran@users.noreply.github.com", "user": {"login": "galran"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "fc8631e2f11d85c84ab9cc711aacb9c589b6f71a", "messageHeadline": "[js/web] Fix conv2dMatmul errors due to #18452 (#18562)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 34, "deletions": 3}}, {"node": {"oid": "b9fd9c5665c998fea8786a2e9fee2776e667845c", "messageHeadline": "remove dead code in openvino EP (#18457)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 2, "deletions": 14}}, {"node": {"oid": "dd355e39a063c124142f60d6cc14f6d48692e1f7", "messageHeadline": "[js/web/training] Added parameters methods (#18250)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 198, "deletions": 40}}, {"node": {"oid": "a2fd8a6fc083f43d6535f5acd24219c140812c87", "messageHeadline": "[QNN EP] Return INVALID_GRAPH if failed to load from context binary (\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 241, "deletions": 209}}, {"node": {"oid": "2f608338cb46398fc3806cb6d1fd3ba7961b1a9f", "messageHeadline": "Setup default python formatter for new python plugin (#18563)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "7b2aefa85688a02a58c5dd7bddc90e7f81f44c3a", "messageHeadline": "undo hipify of __half to rocblas_half (#18573)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "b9c935f6050b3a57e23dbb79e739489f25f6924a", "messageHeadline": "[ROCm] Some fixes in tunable (#18575)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 9, "deletions": 2}}, {"node": {"oid": "62f00ad8e7b7bbaf144e9af2bb19d9bf63dcd291", "messageHeadline": "[CoreML] Add Softmax and Split op support (#18358)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 394, "deletions": 7}}, {"node": {"oid": "6f3c1f9dc9c08ec52c3c2e975e35308b08219494", "messageHeadline": "[ROCm] Update ck for GemmFloat8 (#18487)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 100, "deletions": 87}}, {"node": {"oid": "1c79897c90f959d30ed68c9b36d82be0024d806b", "messageHeadline": "[QNN EP] Support LpNormalization (#18561)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 40, "deletions": 1}}, {"node": {"oid": "43a5147e015e105547aa0e6862462a352fa43c5f", "messageHeadline": "Memory optimization refactor and refinement (#17481)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2639, "deletions": 1040}}, {"node": {"oid": "64dacc2892d31603a5723959d308bb9c4b05d0ea", "messageHeadline": "[js/webgpu] Add BatchNormalization Op (#18468)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 688, "deletions": 0}}, {"node": {"oid": "fa106942a7962e68f1659cd65f5a7cdb498b8c03", "messageHeadline": "[js/webgpu] Refactor matmul conv to support uniforms for matmul (#18452)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 174, "deletions": 98}}, {"node": {"oid": "42c6799c59b5770809a6b4df208d3da5a0270486", "messageHeadline": "Update transpose optimization to be more QDQ aware (#18444)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 959, "deletions": 494}}, {"node": {"oid": "841f7ed3e0c393b22b1631c090c61b20fc62f876", "messageHeadline": "[[JS/Web]Added uniform to Expand op. (#18558)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 49, "deletions": 8}}, {"node": {"oid": "1c555c5fc11d673df9db4f08ebf389c9929e85c0", "messageHeadline": "[JS/Web] Resize & BiasSplitGelu fp16 support (#18536)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 81, "deletions": 75}}, {"node": {"oid": "3f0ebd673622d3663011ae33fc6070f1f2ea3af3", "messageHeadline": "Fix opset import in GemmFloat8 python unit tests (#18489)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "32fabb555501a020751b6123de94c7fc14086f2b", "messageHeadline": "Fix opset version of the optimizer in function generate_artifacts (#1\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 27, "deletions": 1}}, {"node": {"oid": "89723c8612d26d09e0e5995de6f200249035423d", "messageHeadline": "[WebNN EP] Mark and fallback unsupported op for WebNN CPU backend (#1\u2026", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 105, "deletions": 83}}, {"node": {"oid": "3bc9efc7b2ded982995a381e177c388b64d64b1f", "messageHeadline": "[ORTModule] Adjust Attention Patterns for Efficient Attention ATen Fa\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 103, "deletions": 90}}, {"node": {"oid": "7c573054b61bb44e5ee690fbee80aab359b28282", "messageHeadline": "[QDQ Optimizer] Fix logic that drops Q/DQ ops from QDQ split node gro\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 147, "deletions": 32}}, {"node": {"oid": "62da3b1ca43f29b3900d0db5a4a1ee8726a75b3e", "messageHeadline": "SDXL Latent Consistency Model (LCM) optimization (#18526)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 570, "deletions": 148}}, {"node": {"oid": "d455b0f8fd0b0d4bed256fd6089cd20bc9b435b0", "messageHeadline": "[js/web] use Chrome in CI for npm tests (#18522)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 19, "deletions": 55}}, {"node": {"oid": "ac8598a837083dca599ca260152b71d14946de98", "messageHeadline": "[js/webgpu] enable f16 for concat (#18528)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "81a763a9eb559261b79fd3b7d7c36a63c0413fde", "messageHeadline": "Make TensorShapeVector to use InlinedVector<Int64_t> to reduce on tem\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 11, "deletions": 38}}, {"node": {"oid": "680a526e734d497c0280e5ffdf9a738d0e38aeb7", "messageHeadline": "Training packaging pipeline for cuda12 (#18524)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 209, "deletions": 0}}, {"node": {"oid": "2a016225367d7a7ec4bd8b75a3653b0b93b97720", "messageHeadline": "Hide NPU Adapter selection behind macro (#18515)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 18, "deletions": 5}}, {"node": {"oid": "29a409acaa3f8cd8639771c0b4d46d790094aa1c", "messageHeadline": "Add missing flags DISABLE_FLOAT8_TYPES in GemmFloat8 custom operator \u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 204, "deletions": 31}}, {"node": {"oid": "a608c002a3572fdea18817885055014d658b8af6", "messageHeadline": "fix past-kv in general LLM exporter (#18529)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "c7fd930330bd6d557ded5b0f2ca99fe4097d9b29", "messageHeadline": "[js/web] unify resolve rules for \"Clip\" (#18527)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 9, "deletions": 13}}, {"node": {"oid": "abdf8b7c3f6869f781cf21c2918edcf3ce296491", "messageHeadline": "[js/webgpu] Optimize broadcast binary. (#18185)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 24, "deletions": 6}}, {"node": {"oid": "cc542024ce3bd94dfaaabd6100c281cfc4bd2595", "messageHeadline": "Create edges with arg positons correctly accounting for non-existing \u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 217, "deletions": 33}}, {"node": {"oid": "247ce218595acad95a5beeb004cf4c8e74d367d3", "messageHeadline": "[js] optimize eslint config (#18460)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 68, "deletions": 18}}, {"node": {"oid": "1dd9bf53400364d022f3cba7af8c42af06535c30", "messageHeadline": "Remove setup_env_azure.bat (#18482)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 11, "deletions": 13}}, {"node": {"oid": "1af06815540a9a10a6ff5feb3fd8c3f02c95cd77", "messageHeadline": "Bfloat16 support for MatMulBnb4, Training support bitsandbytes>=0.41.\u2026", "author": {"name": "Jambay Kinley", "email": "jambaykinley@microsoft.com", "user": {"login": "jambayk"}}, "additions": 210, "deletions": 47}}, {"node": {"oid": "d97fc1824f3c71e44e40206d920f33bb4c5adb96", "messageHeadline": "Create a new Python Package pipeline for CUDA 12 (#18348)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 242, "deletions": 36}}, {"node": {"oid": "3bcc137eb423ada476118949876611be87636bb4", "messageHeadline": "Tiny change to trigger the update of DORT's CI image (#18507)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "dc9ab4f8213cbef1a0ca93d2630b77fbc13d4da3", "messageHeadline": "Update setup.py: replace libcudart.so.12.0 with libcudart.so.12 (#18501)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "97cc40d75a50e4c10c3f9232bb52fb76db5a7f9b", "messageHeadline": "Add fusion patterns for conformer-transducer model (#18461)", "author": {"name": "Akshay Sonawane", "email": "111780983+apsonawane@users.noreply.github.com", "user": {"login": "apsonawane"}}, "additions": 802, "deletions": 3}}, {"node": {"oid": "53917a33536ab8873264c55c9cac8d91d5a8d040", "messageHeadline": "Move up members in Lite Custom Op hierarchy for possible memleaks. (#\u2026", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 30, "deletions": 17}}, {"node": {"oid": "9364c05170d78c4516886dc91ec86afdce06ad6d", "messageHeadline": "Update web-ci.yml: remove depth=1 (#18500)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "34c54244567af3157a3d37e6d42b9bb918931fbc", "messageHeadline": "[js] update a few packages (#18499)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 10341, "deletions": 64}}, {"node": {"oid": "02333293dec94922585a4aed39bd331128b643a6", "messageHeadline": "Removed all the deprecated python training code and related tests and\u2026", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 21, "deletions": 16534}}, {"node": {"oid": "cbb85b48749a42d6120ac78e40fcc9930814ab37", "messageHeadline": "[CoreML] Adapt to `MLMultiArray.dataPointer` deprecation (#17726)", "author": {"name": "Nicol\u00f2 Lucchesi", "email": "nicolo.lucchesi@gmail.com", "user": {"login": "NickLucche"}}, "additions": 71, "deletions": 36}}, {"node": {"oid": "41f9379f3cd74b1eccf0ef2cbc3bf0ce7c09fc4f", "messageHeadline": "Update NDK version to 26.1.10909125 (#18493)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "fac3e33da510c27c7a2631cf44a79923ee14e09f", "messageHeadline": "[js/web] JSEP Attention & MultiHeadAttention (#17742)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 1866, "deletions": 0}}, {"node": {"oid": "a5537f2f563d4975c7e6121a7eb260bbbfd9455a", "messageHeadline": "[WebNN Ep] Slice's axes and steps inputs should be constant initializ\u2026", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 16, "deletions": 12}}, {"node": {"oid": "1a2946091968fad57e52dd632967a870e0265b06", "messageHeadline": "rope support 4D input tensor (#18454)", "author": {"name": "kailums", "email": "109063327+kailums@users.noreply.github.com", "user": {"login": "kailums"}}, "additions": 103, "deletions": 26}}, {"node": {"oid": "5eb5056c610e274494f182c63c06b30ef0761930", "messageHeadline": "Always run emsdk_env.sh before build.py, even when ccache is disabled\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 20, "deletions": 13}}, {"node": {"oid": "d73073d491d1543fd0fa746bbc8167f85da8488e", "messageHeadline": "remove full protobuf requirement for tensorrt ep (#18413)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "f17b6afe3c5241525c3ee1384f98dbef64bcffbc", "messageHeadline": "[TensorRT EP] Fix bug for no nodes in subgraph at GetCapability (#18449)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 16, "deletions": 6}}, {"node": {"oid": "adb56df2e8de61862c0835c985fb0ba748499b05", "messageHeadline": "Aciddelgado/gqa local (#18375)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 682, "deletions": 537}}, {"node": {"oid": "6a4e4488da75b5b482ef449dfff20309e8b15744", "messageHeadline": "[QNN EP] Support Qnn MatMul with 2 dynamic inputs which are uint16 qu\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 125, "deletions": 8}}, {"node": {"oid": "e7a524fea9599dc4b5e5171cb14c16389b7d58a4", "messageHeadline": "Update to allow large models to be checked for mobile support. (#18357)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 54, "deletions": 11}}, {"node": {"oid": "b6b9aff60846f03b4d68193e2e33afeab8c32c57", "messageHeadline": "Allow empty shapes and do not validate them for inputs/outputs (#18442)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 9, "deletions": 7}}, {"node": {"oid": "3588fbac1377eb2a74fcf82f8d8768c7c00397d3", "messageHeadline": "[TensorRT EP] Fix memory leak for cudnn/cublas  (#18467)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "b291b20fa02b14ad243ef94ce6d72223dbe63ee9", "messageHeadline": "[JS/Web]Added uniforms support to Slice op. (#18422)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 92, "deletions": 22}}, {"node": {"oid": "999752a35d414acc214982b205d16a93768b0699", "messageHeadline": "[WebNN EP] Support GreaterOrEqual and LessOrEqual ops (#18411)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "119e86ec161849b2d904b5423cd6612a4ce31194", "messageHeadline": "SDXL demo: Add Option to disable refiner (#18455)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 37, "deletions": 22}}, {"node": {"oid": "16d7f5519318fc309eccdc957c14eba0bfe5bb10", "messageHeadline": "lora conv1d replacement (#16643)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 282, "deletions": 0}}, {"node": {"oid": "751aa8d31ad10db97762d1a4a60237aae2705a47", "messageHeadline": "fix axis of layernorm for UpstreamReshape (#18425)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 155, "deletions": 169}}, {"node": {"oid": "18a3675bf73f86e05a200428c06c053357bbc51b", "messageHeadline": "[TensorRT EP] Only instantiate TRT builder once  (#18100)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 28, "deletions": 7}}, {"node": {"oid": "6f9f653ada3966e11f454440c066317930d95a5f", "messageHeadline": "[wasm] increase test max memory from 2G to 4G (#18459)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6f863ae2ad5cf86dabe18100cdf3528b0696c1fe", "messageHeadline": "Allow optional axes tensor to be null and ignore it as optional (#18423)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 896, "deletions": 10}}, {"node": {"oid": "cc840c5289e581997ace2f6e414558ca8b886e90", "messageHeadline": "Fix a bug in SaveInputOutputNamesToNodeMapping function (#18456)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "0a4d76d98b8eea5bb863685944577d609c5bac8d", "messageHeadline": "MLAS AArch64 quantized int4 Gemm kernel (#18031)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1744, "deletions": 205}}, {"node": {"oid": "586f06f5a1448e80be19831bb57a23e836af0a33", "messageHeadline": "[js/web] set noUnusedParameters to true and fix a few bugs (#18404)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 146, "deletions": 114}}, {"node": {"oid": "ed89ca573a83e5adcf86a8e2cae912b6eeb9a335", "messageHeadline": "[ORTModule] Support User Config for Triton Codegen, Bugfix for Reduce\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 53, "deletions": 8}}, {"node": {"oid": "b0699d901c8e8e8dcfe4fd3b72d54ef7bef5bf75", "messageHeadline": "Support Graph Input and Initializer for GatherToSplit Fusion (#18412)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 243, "deletions": 45}}, {"node": {"oid": "d738ff16ecf7ba5c21211e745c492153eea15c59", "messageHeadline": "SDXL demo: consistent opt shape and seed (#18445)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 125, "deletions": 56}}, {"node": {"oid": "05526b354b0d29049aa29427def6d861cdec5faa", "messageHeadline": "Adding new yaml file for downloading cuda, and trt from azure blob (#\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 51, "deletions": 0}}, {"node": {"oid": "f9af94009bc9d989848198ae4b759d8c9a4cfea7", "messageHeadline": "onboard MoE (#18279)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 4219, "deletions": 97}}, {"node": {"oid": "27d068569a4f300a3d76fdb1997276a8cffc724b", "messageHeadline": "Remove Node.js tool installer task from web ci pipeline (#18434)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "d22b1af5daf11d10d85d5e95bde5cadc2ab8b7b2", "messageHeadline": "[js/web] add CI steps to log info for test failure investigating (#18\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 44, "deletions": 4}}, {"node": {"oid": "a6b515fad504e394169897eee3bbcc9d82caec79", "messageHeadline": "[QNN EP] Update Where Op UT to include the issue relate to data layou\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "c9d5345c46d51f04edc63ffbdbd93d10e37d9714", "messageHeadline": "[QNN EP] Clean-up todo for OnnxInputInfo (#18416)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 53, "deletions": 54}}, {"node": {"oid": "5aeed6263014327b6621a78ed72c960d6d09ae7e", "messageHeadline": "Bump axios from 1.3.4 to 1.6.1 in /js/node (#18400)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "3e1cf71067575809e6ec6b0b09820c5099fbe832", "messageHeadline": "[TensorRT EP] Fix bug for handling outer scope values in GetCapabilit\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 50, "deletions": 43}}, {"node": {"oid": "a09099f2dd06424a3f3485770198626c303da8d2", "messageHeadline": "Remove XNNPack from web pipelines (#18419)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "0b1618522387734f24fa2f0dc181e78337e7c665", "messageHeadline": "build wasm with linux (#18106)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 12, "deletions": 5}}, {"node": {"oid": "897c1c1f0570ad445b0e3ef8316aad63ffff7daf", "messageHeadline": "Set DML package name correctly in CI (#18405)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "8ff41aea09abf03ac5a9ac9d79534b24fc0ce171", "messageHeadline": "Fix 4 more bad delegates missing the attribute that cause iOS AOT err\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 351, "deletions": 283}}, {"node": {"oid": "37d8bed53d88d6fff85b1fd86827f83e5573ca06", "messageHeadline": "[ROCm] add migraphx into onnxruntime-training-rocm package (#18339)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "f19c67359512657d77df7504f352d0eb29b4914a", "messageHeadline": "If Branch Constant Folding (#18105)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1400, "deletions": 10}}, {"node": {"oid": "a62a500ae1086d884c788a3ed36309e49a70f9e2", "messageHeadline": "[ROCm] Update CK version (#17628)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 30, "deletions": 28}}, {"node": {"oid": "c3b54790562407de404bed31390bae867aa448bf", "messageHeadline": "Remove extra CUDA version flag (#18397)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "4f2bd3862db3a83591f81a29f10b0768a9116b23", "messageHeadline": "[QNN EP] Ensure QDQ Split input/output quant params are equal (#18332)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 358, "deletions": 84}}, {"node": {"oid": "0d22d64420516315e62e26ad6dc9994f376d2031", "messageHeadline": "Update SDXL demo and documents (#18395)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 265, "deletions": 144}}, {"node": {"oid": "949ac4b7ce08bcf4a675859a07d1d8784dad2c28", "messageHeadline": "[js/webgpu] Support uniforms for gather (#18312)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 40, "deletions": 11}}, {"node": {"oid": "4a8203033930da506b356cdaf88b1531d8d8fca3", "messageHeadline": "[ORTModule] Symbolic Shape Support for Triton Codegen (#18317)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 324, "deletions": 120}}, {"node": {"oid": "73ed34ac4b0d6a6717010c5d367fe5675efbfab6", "messageHeadline": "[WebNN EP] Support numThreads option for WebNN CPU device (#18054)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 30, "deletions": 6}}, {"node": {"oid": "cbf0cf06db251594ae3a24877a131b5f75e729d8", "messageHeadline": "[WebNN EP] Disable clamp fusion for WebNN GPU (#18386)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "8d298f6f78f3280ece8d42ddb50caa6e81a6826f", "messageHeadline": "Fix xnnpack compile error on arm32 (#18291)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 13, "deletions": 1}}, {"node": {"oid": "a46c79d211d438e7594147f8725c3c5acd4ef834", "messageHeadline": "fix llama2-70b bug, add document (#18398)", "author": {"name": "Frank Dong", "email": "123416088+frank-dong-ms@users.noreply.github.com", "user": {"login": "frank-dong-ms"}}, "additions": 26, "deletions": 7}}, {"node": {"oid": "646f77a94ba0e128e4d7c1e9290440ab4f4a1bcf", "messageHeadline": "Align context virtuals (#18396)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d87d480857a3645f5c3efd414ddcdf51c0839ebe", "messageHeadline": "Remove deprecated vscode settings (#18349)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 0, "deletions": 19}}, {"node": {"oid": "0c8c0014f6f4a52a2e9783dfd8f12b32f48d382b", "messageHeadline": "[js/webgpu] Use builtin num_workgroups to fix shader key conflict (#1\u2026", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 24, "deletions": 23}}, {"node": {"oid": "6b0c97b43f8817bef31a5e0200c800b567e44794", "messageHeadline": "[js/web] fix typescript type check (#18343)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 29, "deletions": 28}}, {"node": {"oid": "8dba6efd612ee647923670ec0919d8192d815cee", "messageHeadline": "[js/webgpu] Add uniforms support to concat op (#18238)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 42, "deletions": 12}}, {"node": {"oid": "64c91d790bbb3e2f5a97ec6fb11650cd71644cb6", "messageHeadline": "Fix ability to use patch on Windows CI machines (#18356)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 20, "deletions": 42}}, {"node": {"oid": "28c23aed041010726eea9ef7ed58528e98701b8c", "messageHeadline": "[js/webgpu] Fix conv2d with activation (#18388)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 41, "deletions": 55}}, {"node": {"oid": "2d23b4e117c34094cd40e70136d21998ff0a9fee", "messageHeadline": "Update min macos version (#18251)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "87744e55faa8c7570d17f29f54f2ffdb8b702d04", "messageHeadline": "fix reference to Microsoft.GSL::GSL in CMake build scripts when enabl\u2026", "author": {"name": "Bart Verhagen", "email": "bverhagen@users.noreply.github.com", "user": {"login": "bverhagen"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "dd1bb760eb4e0c73363633f9798f0662f90b1638", "messageHeadline": "[js/webgpu] Fix scalar uniform (#18318)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 14, "deletions": 13}}, {"node": {"oid": "d9558857913b32a9d701d217362017b892a2769c", "messageHeadline": "Update stale.yml to fix start-date bug (#18376)", "author": {"name": "sophies927", "email": "107952697+sophies927@users.noreply.github.com", "user": {"login": "sophies927"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "59262dfc63a68d8b33d25ae0b855297dae6bac0f", "messageHeadline": "Add cuda context headers to zip (#18330)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 27, "deletions": 4}}, {"node": {"oid": "1ff894898a6ca84accee4c8b1b9eccaf2c345332", "messageHeadline": "Bump actions/stale from 4.1.1 to 8.0.0 (#18149)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "829d8023378c9b9890347f51a838a1e3664bcd88", "messageHeadline": "[js/webgpu] Support uniform for softmax (#18345)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 46, "deletions": 18}}, {"node": {"oid": "f237b0b1f855ebcfcba093b5e684fd3069fd51e6", "messageHeadline": "[QNN EP/Quantization] Add MinimumRealRange extra option to quantizati\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 141, "deletions": 4}}, {"node": {"oid": "25fbc2b0abb18a1277c0f5d4b6c38bf58b9de154", "messageHeadline": "fix fused relu activation (#18303)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "2c22b498762bb5788e119fc7da3a72f12f35631b", "messageHeadline": "Fix rust compile issues and add GH action to run build validations an\u2026", "author": {"name": "David Justice", "email": "devigned@users.noreply.github.com", "user": {"login": "devigned"}}, "additions": 244, "deletions": 58}}, {"node": {"oid": "8d503138168d4ae54869c82b1032a38c6cb9f3a7", "messageHeadline": "[Migraphx EP] Static int8 QDQ support (#17931)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 671, "deletions": 61}}, {"node": {"oid": "55c19d6ab5d49a95f58e66baccdb5f81092e4eeb", "messageHeadline": "[QNN EP] Enable option to set QNN context priority (#18315)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 131, "deletions": 29}}, {"node": {"oid": "7a3da4526f98c9cfc6387a5faa1edeec7d88ef17", "messageHeadline": "add bfloat16 support for CUDA Neg kernel (#18306)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 16, "deletions": 6}}, {"node": {"oid": "4dc63692f83ad1217e0964f6efbf78f4c83dfb60", "messageHeadline": "Add FlattenAndUnpad Op (#17845)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 448, "deletions": 118}}, {"node": {"oid": "885bf3561da143ad90d30b1d5959a9b065f5a6c9", "messageHeadline": "Add tool to fix lines > 120 chars. (#18293)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 136, "deletions": 2}}, {"node": {"oid": "c250540722a8028250de1d572365ec73ef6ecedf", "messageHeadline": "Bump linter versions (#18341)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 102, "deletions": 78}}, {"node": {"oid": "812532592e070823f0362af8c3219ca6b01eda77", "messageHeadline": "Add a build validation for Linux ARM64 cross-compile (#18200)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 104, "deletions": 17}}, {"node": {"oid": "68fab24c22e1453ebfcd9d67e857c6531985faa0", "messageHeadline": "Update stale.yml (#18304)", "author": {"name": "sophies927", "email": "107952697+sophies927@users.noreply.github.com", "user": {"login": "sophies927"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "a37e6a503b3073605b1eb765d1b0ff5a61360991", "messageHeadline": "Update Abseil raw_flat_hash visualization (#18329)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 38, "deletions": 12}}, {"node": {"oid": "a0eeeafa8050bc354fae279138147f3ea59f83eb", "messageHeadline": "[QNN EP] Session option for graph optimization (#18262)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 270, "deletions": 37}}, {"node": {"oid": "c8def0cc5193783492c0ed153f0d71527a80857e", "messageHeadline": "Add LLaMA GQA ragged batching (#18337)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 257, "deletions": 106}}, {"node": {"oid": "34f77eaa243ed16bbcea8fa585c9f89539488b27", "messageHeadline": "bfloat16 support for quickgelugrad (#18336)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 11, "deletions": 4}}, {"node": {"oid": "2151c79bf190cdd71794d224f8efea4f8b9cd207", "messageHeadline": "Tune ORTModule logging experience a bit (#18298)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 101, "deletions": 24}}, {"node": {"oid": "8044e5f60332940217576a067d99b454a83c5457", "messageHeadline": "SDXL: Update demo with dynamic shape serving with CUDA EP (#18340)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 107, "deletions": 30}}, {"node": {"oid": "3dece27f51f60f49f899a0321f019539168679ce", "messageHeadline": "GQA Flash Attention with Attention Mask (#18283)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 1049, "deletions": 357}}, {"node": {"oid": "10df847bafcd53221adf5cec81827633275a4583", "messageHeadline": "[js] fix linter out-of-memory issue (#18307)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d117a8010f25393af6d6819ecc59121443eb9161", "messageHeadline": "fix typo (node)->(browser) in linux-wasm-ci.yml (#18309)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "096307c64b789354e2c20e915e918e66d911dd3e", "messageHeadline": "Do not run AOT function inlining when the model does not define any l\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 22, "deletions": 3}}, {"node": {"oid": "606356d0b1bb5aa8a4cfe14f3263efafba5051eb", "messageHeadline": "[js/webgpu] Simplify the Resize shader when noScale is true (#18321)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "6127dd1d2d802f0f9e615cac399d6d84356083aa", "messageHeadline": "implement gridsample 20 (#17744)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 1383, "deletions": 158}}, {"node": {"oid": "83c0275354bf21845df21d83dc9e4e249c8e74ed", "messageHeadline": "add bfloat16 support for ConcatTraining and SplitTraining ops (#18280)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a16d528399a581a8d85de109e335897487636516", "messageHeadline": "[JS/Web] Added Uniforms support to binary ops. (#18260)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 40, "deletions": 43}}, {"node": {"oid": "800ae7742cbc09d79753070f96fe933da4e23e6c", "messageHeadline": "[DML EP] Add RotaryEmbedding (#18158)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 461, "deletions": 8}}, {"node": {"oid": "9868a7137398608ef930755e8a96e3bda84da755", "messageHeadline": "[Fix] Stages to Run couldn't be selected (#18310)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "4f15b427286387313503a39489bd1d969dfab84a", "messageHeadline": "Customize _get_tensor_rank for model export in stage3 (#18294)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 62, "deletions": 5}}, {"node": {"oid": "630c877b43bb6f13271afb52e8fd5e008314028d", "messageHeadline": "Zhijxu/improve ortmodule python perf a little bit (#13716)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "00c2bf39bdbf5096cc9fbcb7bfd803a13d00f35a", "messageHeadline": "SkipGroupNorm fusion and SDXL Pipeline Update (#18273)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 447, "deletions": 76}}, {"node": {"oid": "276918d93b24b8b845122de5f61d2e091d608cc8", "messageHeadline": "Allow SkipLayerNorm and LayerNorm in rotary attention fusion (#18288)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "fb6737e8934567a4394c02fbd968ae4a48d38d63", "messageHeadline": "Distributed Squeeze and Distributed Unsqueeze (#18269)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 596, "deletions": 8}}, {"node": {"oid": "ad34c67a444d15132fce1c85612f16db7d6fa74e", "messageHeadline": "[QNN EP] Enable Expand op (#18234)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 609, "deletions": 347}}, {"node": {"oid": "3b63d85c253c50099c70ba0db6c141b842bc7cda", "messageHeadline": "Fix unit test when TVM EP is enabled (#18189)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "398ef677bafeff9bd6a76bb03b0517a8a515ad4d", "messageHeadline": "Update protobuf python package's version (#18203)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "b7b8b5b2ce80edb33990c7ae0fedac6ae3c623f4", "messageHeadline": "Fix Eigen-3.4.0 URL and hash (#18290)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "d652b1fe48d1b9874575646a7680aab1e490ec0a", "messageHeadline": "[VitisAI] fix tensor has multi data type (#18188)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "dfafcb58aa5fb262b4ff964ccdcc3271672032c0", "messageHeadline": "[TensorRT EP] Properly set CUDA_INCLUDE_DIR for onnx-tensorrt  (#18274)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "08eaa1c55d3f1c9fa692deb379b4c8cbf9105711", "messageHeadline": "Remove internal enforce for IO binding inputs (#18266)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "84bdf04b255661e8446ded49edda4dd8d730101f", "messageHeadline": "[TensorRT EP] Fix bug for shape tensor input (#18253)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 162, "deletions": 2}}, {"node": {"oid": "26b396418d793cf23db37b643ca833063e9368b6", "messageHeadline": "Block-wise 4b quantization matmul operator change (#18172)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 482, "deletions": 444}}, {"node": {"oid": "2ec1f94bfdeba44f48822197ffdb06f817edf412", "messageHeadline": "Make MlasTestFixture::mlas_tester an inline variable. (#18263)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 254}}, {"node": {"oid": "4c4d79a6120737e093fcd4ce79006e1f928b7816", "messageHeadline": "Change a bitwise logical xor to logical wise (#18246)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "192caee81f8505653b1616d899ac01beb641ee90", "messageHeadline": "Fix Signed Mismatch (#18258)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e207060ac9e1beaa340ce023f018dc792c8527b5", "messageHeadline": "[JS/Web] Added Unifroms support to unary ops. (#18223)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "90f205e79c3ff911a333a06abe86e052fb4f9d05", "messageHeadline": "[QNN EP] Fix Pad UT (#17982)", "author": {"name": "winskuo-quic", "email": "143469905+winskuo-quic@users.noreply.github.com", "user": {"login": "winskuo-quic"}}, "additions": 44, "deletions": 13}}, {"node": {"oid": "c352e9b1f9f2faacbe9551c5ba7c53e77efadc8b", "messageHeadline": "Rework/cleanup the C# build infrastructure for nuget packages. (#18127)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 178, "deletions": 211}}, {"node": {"oid": "4f2096be38bb04b103c2577d5f132c92419b26ad", "messageHeadline": "Update XNNPACK to latest version (#18038)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 794, "deletions": 517}}, {"node": {"oid": "e36d0037654122264ab6ce48d5bf1f282543c51a", "messageHeadline": "Introduce new optimizer Pad + Conv/MaxPool (#18136)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 315, "deletions": 0}}, {"node": {"oid": "016b75260bf292b7607d9a8520ddaab513a3a8b8", "messageHeadline": "Pre-link when creating static library for apple framework (#18241)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 69, "deletions": 24}}, {"node": {"oid": "1439da36fe10df74dcaaf165d08999b15cd955a0", "messageHeadline": "Partially disable QGemm tests for float 8 types (#18196)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 3, "deletions": 18}}, {"node": {"oid": "9f5a6856fe04e64e9576657b3cfdc4ddac17682e", "messageHeadline": "Rerun the flaky ort-web tests automatically (#18187)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "d8d79521ca2b266e631ac0ba7036a682ebb58b5b", "messageHeadline": "Disable ccache for DML (#18230)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8d48d3e9cce4a7b5640051899f824c69a6a67c62", "messageHeadline": "[js/web] optimize reduce related operators (#17957)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 380, "deletions": 11}}, {"node": {"oid": "8978bdc59dedc95ddf8357b4acfb122ec2547ece", "messageHeadline": "add bfloat16 support for where operator (#18118)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "c8e1038eab6c07392fb0891d0fec45e43f1a7487", "messageHeadline": "Optimize 4bit Qlora training (#18131)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 297, "deletions": 67}}, {"node": {"oid": "e3b043ba1790b8a6710829b390ce912675496929", "messageHeadline": "[js/web/training] runTrainStep implementation (#18006)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 441, "deletions": 92}}, {"node": {"oid": "f9d5705db488266bfe51acbf22e78fc95edc7596", "messageHeadline": "[EP Perf] Fix sort (#18174)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "dabd395fdfdb3c5edf91d3b515bab00744b63c60", "messageHeadline": "llama 70b model fusion and shardding (#18175)", "author": {"name": "Frank Dong", "email": "123416088+frank-dong-ms@users.noreply.github.com", "user": {"login": "frank-dong-ms"}}, "additions": 1259, "deletions": 252}}, {"node": {"oid": "178f7caaebf67bcddee8fae8836656613290b8bc", "messageHeadline": "GQA Memory Efficient Kernel (#17920)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 843, "deletions": 312}}, {"node": {"oid": "a2e9ba72d5a5f61e1324ffc2a80d748d01be9120", "messageHeadline": "[JS/Web]Added FusedConv. (#17766)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 339, "deletions": 98}}, {"node": {"oid": "9e8ad398479d9c2dc0ca91a8df89e452d059f6ee", "messageHeadline": "Distributed Reduction (#18206)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 638, "deletions": 108}}, {"node": {"oid": "d87216bcb13c8a3937a74b1cd2160aeb7d9cffb7", "messageHeadline": "Openvino ep ort 23.1 (#17911)", "author": {"name": "Preetha Veeramalai", "email": "preetha.veeramalai@intel.com", "user": {"login": "preetha-intel"}}, "additions": 564, "deletions": 358}}, {"node": {"oid": "69f029797d24e44c6854d9c68231bef174e627e4", "messageHeadline": "[QNN EP] Fix Batch Normalization Op Builder  (#17981)", "author": {"name": "weischan-quic", "email": "138087696+weischan-quic@users.noreply.github.com", "user": {"login": "weischan-quic"}}, "additions": 583, "deletions": 24}}, {"node": {"oid": "819b5a3eba85cca9276c9d763c814eb45067b280", "messageHeadline": "Split KV on MHA and Attention ops (#18007)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 90, "deletions": 33}}, {"node": {"oid": "c181159783d1245adb9bb1af18a469ad7d89df45", "messageHeadline": "[WebNN EP] Restore to use deviceType enum (#18154)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 1, "deletions": 5}}, {"node": {"oid": "d1b85f5fb4fff6fc674e50e2053039c7ded4969e", "messageHeadline": "Reduce LLaMA memory usage (#18181)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 248, "deletions": 186}}, {"node": {"oid": "2b95e74fa113ec168a79974987b2c6b98cecf700", "messageHeadline": "Versioning for custom op (#18088)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 148, "deletions": 37}}, {"node": {"oid": "62c7894ffe15efb7d43d891a326c2cbdcfbb529d", "messageHeadline": "Add mobile CIs to list run by script for external PRs. (#18094)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "ed41a2836c7963f4c46a073ea7bc29f971e06618", "messageHeadline": "Fix cast removal bug (#17953)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 133, "deletions": 18}}, {"node": {"oid": "20f2dd8b6ba1c461f9a8d90a578178eab1ff20f7", "messageHeadline": "use onnx rel-1.15.0, update cgman, cmake/external and requirement has\u2026", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 8, "deletions": 18}}, {"node": {"oid": "95f053c652ac1e7afb3d795ea5df2b165b2429c2", "messageHeadline": "[CUDA] Update GroupNorm and Add SkipGroupNorm (#18091)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1531, "deletions": 241}}, {"node": {"oid": "29e40987e3bc5b8f37c14c8ea6b6eccb620dee42", "messageHeadline": "Update batch file to set PATH for Cuda with TRT (#18182)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 14, "deletions": 4}}, {"node": {"oid": "1c25fe55800b05ffd2d89d2c6f69b39820bc5c4f", "messageHeadline": "Fix PoliCheck (#18180)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "08dce542665598cf659447555f4e0eb6d96bf86a", "messageHeadline": "Improve tunable verbose log (#17328)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 35, "deletions": 35}}, {"node": {"oid": "8a574b874cf5093b9a1e5383046f6e71d9ada4cf", "messageHeadline": "Update setup_env_cuda.bat (#18176)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 9, "deletions": 7}}, {"node": {"oid": "8ed9bd6eca143c791c56c055b7a9b0755040a5f3", "messageHeadline": "Add one more MHA mask pattern (#18164)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 16, "deletions": 0}}, {"node": {"oid": "efef6407bcdd2464f7f38b09579fc6aeb331655e", "messageHeadline": "[ROCm] update rocm package exclude libs (#18130)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 11, "deletions": 8}}, {"node": {"oid": "785e2b1eaec2b2e69267cc9320f7749315c0e686", "messageHeadline": "[js/webgpu] Optimize softmax by vector (#18153)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 30, "deletions": 14}}, {"node": {"oid": "90d1f537cb32f400f5abd16248c7abe1a2738484", "messageHeadline": "optimize SLN with large dimension (#18138)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 45, "deletions": 37}}, {"node": {"oid": "348a963238a5ebd242259ffb9e158034cdd052af", "messageHeadline": "[DML EP] Handle non-raw data in dynamic graph compilation (#18160)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 35, "deletions": 24}}, {"node": {"oid": "4819fbf31c39ad08d1406651317503fb1973d107", "messageHeadline": "Augment blockwise quantization (#18101)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1023, "deletions": 0}}, {"node": {"oid": "be2f72a315637c68d92152744ff482a20fd2db9c", "messageHeadline": "[QNN EP] Disable early termination in GetCapability (#18140)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 6}}, {"node": {"oid": "9bba99087185d2e8fe6feb007033983b83ea01fe", "messageHeadline": "[js/web] fix a few package consuming problems (#18109)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 42, "deletions": 76}}, {"node": {"oid": "436056dcd7c0533b52325fd052ba9fd219aacd66", "messageHeadline": "Revert \"Disable dml stage in windows GPU pipeline temporarily. (#1803\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 17, "deletions": 0}}, {"node": {"oid": "8ebdd3bbcafc7e195e41b2aa3881d258d365f608", "messageHeadline": "Fix regression in perf test runner (#18139)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0e34100484594508e810ce43ce242112edac6444", "messageHeadline": "create memory descriptors based on the tensor dimensions (#15848)", "author": {"name": "snadampal", "email": "87143774+snadampal@users.noreply.github.com", "user": {"login": "snadampal"}}, "additions": 22, "deletions": 3}}, {"node": {"oid": "24f9c1afe3a1972cfacd9b4bcd41f2227d369bf0", "messageHeadline": "Distributed Expand (#18126)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 470, "deletions": 0}}, {"node": {"oid": "8daabf3f15329e58d28557c9fdfd1254836b6573", "messageHeadline": "Tune min version supporint custom op ComputeV2 (#18134)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d9695dea6df183f4290d6867f8fae0519f542d83", "messageHeadline": "[WebNN EP] Remove Conv initializer constraint for GPU  (#18129)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 15, "deletions": 13}}, {"node": {"oid": "28ad3ff799e163bf5c854db03e87086a61e14256", "messageHeadline": "Fix stale bot issue (#18064)", "author": {"name": "sophies927", "email": "107952697+sophies927@users.noreply.github.com", "user": {"login": "sophies927"}}, "additions": 34, "deletions": 22}}, {"node": {"oid": "2eeafc37bca21dc8bf337dda7020b486543162d7", "messageHeadline": "Enable global TRT timing cache (#17865)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 139, "deletions": 479}}, {"node": {"oid": "58f1d15d19006464546c73ac6fbed95ff5c90b0a", "messageHeadline": "Replace Transpose with Replace if they are equivalent (#18096)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 190, "deletions": 0}}, {"node": {"oid": "b5f242e9789b2add38578ec922c2b7f8cab254a8", "messageHeadline": "GemmFloat8 as a contrib ops (#16051)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1139, "deletions": 0}}, {"node": {"oid": "c10b83eb68e64d66a836b63c6976ecfdd58257bc", "messageHeadline": "Update python cryptography version to 41.0.4 (#18056)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9c323106735535b6dab6b476648faac0ad185e21", "messageHeadline": "Distributed Reshape Implementation (#18068)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1870, "deletions": 40}}, {"node": {"oid": "b79ea7481930ae09980f61982a8b7b19303a0eca", "messageHeadline": "Add updates to LLaMA scripts (#18076)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 59, "deletions": 38}}, {"node": {"oid": "0f3a067d3a7d42bbc860b97d7ed39e5d7cdd5d47", "messageHeadline": "[FIX] reorder initializer (#18097)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b7408f73892b99f0eab1256b7e8ee4013b57d90c", "messageHeadline": "[ORTModule] ATen Efficient Attention and Triton Flash Attention (#17959)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 2037, "deletions": 93}}, {"node": {"oid": "37873be86df6715256b79818c1e7759848bf5c35", "messageHeadline": "enable reduce ops on opset18 (#18053)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 275, "deletions": 527}}, {"node": {"oid": "52f496835949537fe359a544ce14ec4810860b6d", "messageHeadline": "[js/webgpu] Change timestamp-query-in-passes to timestamp-query (#18108)", "author": {"name": "Yang Gu", "email": "yang.gu@intel.com", "user": {"login": "gyagp"}}, "additions": 50, "deletions": 48}}, {"node": {"oid": "b7bee621cd983926ed801b420c253cd26ea8e6ce", "messageHeadline": "[CUDA] Remove shape warnings in NHWC <> NCHW unit tests (#17992)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "7c18c60bc20dcf2006b9e546e3162bdc7e18823c", "messageHeadline": "Change cuda image for tensorRT to the one with cudnn8 (#18102)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 7, "deletions": 23}}, {"node": {"oid": "f2e19a8ccf90e089b47b01ac59a83e6991d53776", "messageHeadline": "Updates to training pipelines to reduce CI time (#18116)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 8, "deletions": 14}}, {"node": {"oid": "a514a6877015bddaa6b98e272720ac9927629e14", "messageHeadline": "Support per-tensor device mesh at op level (#18025)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 176, "deletions": 40}}, {"node": {"oid": "455a9ce61418cca1674fd20222c6bf4c6bf4e3b4", "messageHeadline": "[TensorRT EP] Use latest onnx-tensorrt parser (#18067)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "b023de0bfc7acb2404dfdcc4adc060b7b72fdaa1", "messageHeadline": "Redo #18044 Install CUDA 12.2 on Windows (#18093)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "64de71c5e2af8ee67ce79a9bb8d7244a8238064c", "messageHeadline": "[js/web/training] Add CreateTrainingSession (#17891)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 399, "deletions": 12}}, {"node": {"oid": "0f72739b6db129373d221483d61d6637ec11fb28", "messageHeadline": "Disable ccache for WinML build (#18104)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "538e97cbda5e8c6c7c3a8796ce483da185f2e91c", "messageHeadline": "[DML EP] Add dynamic graph compilation (#17876)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 1127, "deletions": 143}}, {"node": {"oid": "d30d4d372a33640aed78eb4bfe05db31e36f6e2a", "messageHeadline": "Add MatMul FP4 and NF4 Support (#18066)", "author": {"name": "Jambay Kinley", "email": "jambaykinley@microsoft.com", "user": {"login": "jambayk"}}, "additions": 2236, "deletions": 0}}, {"node": {"oid": "d88d52eeadcdc2813e478dbb2d8dc6f5575258d6", "messageHeadline": "[aarch64] Remove mmla kernel support from apple (#18082)", "author": {"name": "snadampal", "email": "87143774+snadampal@users.noreply.github.com", "user": {"login": "snadampal"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "706e13e0c95a730181bca62c348d3283a9194e11", "messageHeadline": "implement affinegrid cpu kernel (#17777)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 457, "deletions": 0}}, {"node": {"oid": "2c6b31c5aa05bdce26ccd1af58bb194f880166ed", "messageHeadline": "FP16 optimizer automatically detect DeepSpeed compatibility (#18084)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 223, "deletions": 31}}, {"node": {"oid": "ae8561979f494029c863dafb67bae05639ebff60", "messageHeadline": "Introduce new optimizer MatMul + BatchNormalization (#17915)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 543, "deletions": 9}}, {"node": {"oid": "76e275baf44d5bd882fd298d3b86d824eb113435", "messageHeadline": "Merge Cuda docker files into a single one (#18020)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 90, "deletions": 730}}, {"node": {"oid": "6ec45f2ba590fabad99159a44fd6e48a5a9b03f0", "messageHeadline": "Merge aiinfra-linux-ARM64-CPU-2019 and onnxruntime-linux-ARM64-CPU-20\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "efa0cc2562c28e6376717b46ebc83dd29b68d348", "messageHeadline": "implement isinf20 and isnan20 (#17874)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 389, "deletions": 101}}, {"node": {"oid": "abb329179adae0029ef492c251984fcfd78224c4", "messageHeadline": "Update win-wasm-ci.yml: increase the timeout value (#18023)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e63ccd3cbb9e2479af94a69a0e2c9bb9b59a54e4", "messageHeadline": "Install CUDA 12.2 on Windows (#18044)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 36, "deletions": 22}}, {"node": {"oid": "eb47008049a7aa0b617340bf2372723d0e873752", "messageHeadline": "[js/webgpu] FP16 Cast, Resize (#18035)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 24, "deletions": 25}}, {"node": {"oid": "688524a9abca24d576297224cd48f92f283b0f11", "messageHeadline": "[CUDA EP] Add warning logs when adding memcpy nodes (#18032)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 30, "deletions": 10}}, {"node": {"oid": "555b2af7d67d56b3bd8b51dc375d7fd144c761d7", "messageHeadline": "[TensorRT EP] Add unit test for user provided cuda stream (#17974)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 126, "deletions": 9}}, {"node": {"oid": "4ffd022b0b1a62880d2498679f360878fdbbd796", "messageHeadline": "[TensorRT EP] Refactor of TRT plugins support (#17946)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 88, "deletions": 32}}, {"node": {"oid": "2c50b75a26429ef3146d1c6c541f3a3112aa7c83", "messageHeadline": "Functions Ahead Of Time inlininng (#17764)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 450, "deletions": 138}}, {"node": {"oid": "f3cfe08c422be8671883fd3ae8617f05f22adb26", "messageHeadline": "[JS/Web] Enabled 1d spacial input to GlobalAveragePool (#17973)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 23, "deletions": 27}}, {"node": {"oid": "780ee186d72b74a77c4a2752d7a84c3807206ee6", "messageHeadline": "[aarch64] Implement QGEMM kernels with UMMLA/SMMLA instructions (#17160)", "author": {"name": "snadampal", "email": "87143774+snadampal@users.noreply.github.com", "user": {"login": "snadampal"}}, "additions": 3833, "deletions": 0}}, {"node": {"oid": "2a17d5cf32900fa0100959eace6e303c82f86bdc", "messageHeadline": "LLaMA Model Optimization (#18021)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 5897, "deletions": 563}}, {"node": {"oid": "8a12b2cea6c80f312045f4ac74b818cb5b53fa35", "messageHeadline": "[js/webgpu] Fix the transpose error when dims > 4D (#18027)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 59, "deletions": 25}}, {"node": {"oid": "f0d5ea5930bee7e87c1d93e14d4d28c8af3c8cde", "messageHeadline": "[QNN EP] Disable flaky test QnnCPUBackendTests.MatMulOp_Broadcast (#1\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "b7ae293be05c89a0cb623feec4d2d2cbf006e4c3", "messageHeadline": "Support large model export using multi-gpu (#17990)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 385, "deletions": 0}}, {"node": {"oid": "444a0eda309e0fadf51c63790b6da78258f96a10", "messageHeadline": "Avoid one time clone to save memory peak (#17934)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 32, "deletions": 23}}, {"node": {"oid": "009cd4ea2e0621459806010cea7d7533d0acb39d", "messageHeadline": "Allow cuda custom ops allocate deferred cpu mem (#17893)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 100, "deletions": 17}}, {"node": {"oid": "2f57625cb01300b538bce61ea51caffa236b4732", "messageHeadline": "[TensorRT EP] Add stream sync after enqueue (#18026)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "020824ed509893b13aaf6fdc3e651c7a341f7273", "messageHeadline": "Update ONNX to 1.15.0rc1 (#17914)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 73, "deletions": 8}}, {"node": {"oid": "a43c57f59db35d8cceef1ff8f44985d745eb94b4", "messageHeadline": "ResizeGrad CUDA/ROCM kernel implementation (#17772)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 605, "deletions": 25}}, {"node": {"oid": "cc7e8cc21f83df3a41d9736dba9211bb832764ad", "messageHeadline": "Update dockerfiles/Dockerfile.source to avoid installing onnx (#17975)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "99b8dcaae2ac033b10880bc5bc7b0e89b37fe466", "messageHeadline": "Disable dml stage in windows GPU pipeline temporarily. (#18034)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 0, "deletions": 17}}, {"node": {"oid": "35ecce45496ee752bc3f85618eba713bb50e6069", "messageHeadline": "[QNN EP] Reduce overhead of QNN context binary loading (#17965)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 74, "deletions": 80}}, {"node": {"oid": "cbb0e0f83c8253c09041fea342edafea9a2e93fe", "messageHeadline": "Create a new Dockerfile for cuda 12 and trt 8.6.1.6-1.cuda12.0 (#18000)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 190, "deletions": 5}}, {"node": {"oid": "a2c62832748184405ab4ddd341a166ffb1f2fafa", "messageHeadline": "Fix Packed MultiHead Attention (#17996)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 56, "deletions": 44}}, {"node": {"oid": "22947109f2747097e79bf253c4555e7e9984b4bd", "messageHeadline": "[js/web] FP16 LayerNorm, InstanceNorm, SkipLayerNorm (#17630)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 344, "deletions": 190}}, {"node": {"oid": "f9694c5b973605f067f9dbe13c04ae140602bfd9", "messageHeadline": "Bump @babel/traverse from 7.18.5 to 7.23.2 in /js/react_native/e2e (#\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 117, "deletions": 75}}, {"node": {"oid": "59ae3fdfdc77aab5022c7359c7e2e988ed221936", "messageHeadline": "[CUDA] StableDiffusion XL demo with CUDA EP (#17997)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 379, "deletions": 196}}, {"node": {"oid": "61f1a16265a31cf7396adfefe20835a833fcbc13", "messageHeadline": "Fix MHA shape inference (#18009)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 3, "deletions": 9}}, {"node": {"oid": "65575389f4d54ee9f64443160de91d092b8dff90", "messageHeadline": "[DML EP] Enable more MHA masks (#17882)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 26, "deletions": 4}}, {"node": {"oid": "2ef7abf1ff98d3cccbc0bf23661fe228a2357a3b", "messageHeadline": "support for fp16 in GetClipMinMax() (#17967)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 32, "deletions": 17}}, {"node": {"oid": "3456831413494b1bb6f3839414dfc4af8131dea4", "messageHeadline": "[java] Make the backing byte buffer in an OrtValue accessible (#16578)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 141, "deletions": 5}}, {"node": {"oid": "57c87365968050850903eefd4e528b3bcd8abbad", "messageHeadline": "Move a nodejs test to a different machine pool (#17970)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "9356986730a0fd5daab1c09a9de481edd4f4ca8c", "messageHeadline": "Fix AMD builds and enable testing NHWC CUDA ops in one GPU CI (#17972)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 22, "deletions": 11}}, {"node": {"oid": "6832b688a2848f64d4d43ac0084c6609aa0203bc", "messageHeadline": "Fix missing attribute on C# DOrtGetResizedStringTensorElementBuffer d\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 32, "deletions": 21}}, {"node": {"oid": "83547d30672ca1fa11327420037a1be2a6448950", "messageHeadline": "[CUDA] Fix SkipLayerNorm vectorized kernel out-of-bounds read (#17943)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 72, "deletions": 49}}, {"node": {"oid": "cf974f0905a78280c1a47b519e7d6c0b0a6ca069", "messageHeadline": "Bump @babel/traverse from 7.18.2 to 7.23.2 in /js/react_native (#17962)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 113, "deletions": 47}}, {"node": {"oid": "7c17e33c07a36ce0d14b93ae0a74f3774d6a8630", "messageHeadline": "Make CUDA a NHWC EP (#17200)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 1198, "deletions": 287}}, {"node": {"oid": "8abaa7b753c2e77bd344330bf9c55765117a274d", "messageHeadline": "[TensorRT EP] Fix cmake install (#17923)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "ad817d0efa837b18b0c726a467e617133b813fd5", "messageHeadline": "[js/web] optimize tsc for web: split out \"npm prepare\" (#17955)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "f7341e810385727c4fc3dad0c307b745f1479500", "messageHeadline": "enable training for win-wasm-ci.yml (#17954)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 20, "deletions": 5}}, {"node": {"oid": "209b6dbd975efbc792b5ca9ae1dd74b828559148", "messageHeadline": "add onnxruntime_float16.h into header validation list. (#17935)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "ae211999dd606af8ed04e14ea51dca2ec4b45ec7", "messageHeadline": "Attempt to make the usage of the Android emulator in CIs more robust \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 287, "deletions": 193}}, {"node": {"oid": "a55b2688b6d0f88620c4fb0f6af322f62b67b369", "messageHeadline": "Add save_attribute option to quantize_static (#17945)", "author": {"name": "Zhipeng Han", "email": "zhipeng.han@outlook.com", "user": {"login": "zhipenghan"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "11af34440a3e216903fcc9dd7e169969fdc24d03", "messageHeadline": "Add MatMul 4bits support on GPU (#17890)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2389, "deletions": 3}}, {"node": {"oid": "3b69d9b8862417061c34743bf199af805f139316", "messageHeadline": "[QNN EP] Fix index-out-of-bounds bug in Slice builder when initialize\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 176, "deletions": 109}}, {"node": {"oid": "9c65d5558c9a2873811ccc5269a02fc797cae0b2", "messageHeadline": "[WebNN EP] Fixed wasm heap overflow for big model (#17902)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 16, "deletions": 10}}, {"node": {"oid": "c6c3555d0e80b0ce6ecf46c878a3f28424c37a5c", "messageHeadline": "Custom op shape inference API (#17737)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 771, "deletions": 25}}, {"node": {"oid": "762703e037ca1ba9dd2024c42732d45d38f7f4d5", "messageHeadline": "Support output cross qk, dtw and more for whisper model (#17500)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 2561, "deletions": 312}}, {"node": {"oid": "c695de91ee1fb51c426c6b26a587ed961f70253d", "messageHeadline": "Update eval_squad to use API of latest optimum (#17918)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 48, "deletions": 25}}, {"node": {"oid": "7551dd039f67e1adc9fa6d2b2627c6c7f17a0a8d", "messageHeadline": "Fix build break with cuda 12.2 (#17922)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "28c1944561cbd21256368943e452ac304fb8f582", "messageHeadline": "[TensorRT EP] Add back creation cudnn and cublas using external cuda \u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "635d3faa3b3908d2806d009dc6872152cfcfcdda", "messageHeadline": "Remove half of the compiled shaders for gridsample with unused types \u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 6950, "deletions": 106925}}, {"node": {"oid": "67d7eb3ac5cadda092f7d4682875cde475b2fc14", "messageHeadline": "[CUDA] Fix SkipLayerNorm strict mode when skip has broadcast (#17896)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 171, "deletions": 148}}, {"node": {"oid": "4e2a03d5faa60b1aa4049567147e93187aa38a92", "messageHeadline": "[QNN EP] Fix topological node unit traversal during validation (#17913)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 17, "deletions": 2}}, {"node": {"oid": "dad70ad4e80fb6461925d33f331d26ff9c52bb6c", "messageHeadline": "[QNN EP] Handle rank 3 InstanceNormalization with N != 1 (#17897)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 200, "deletions": 15}}, {"node": {"oid": "0c5b1598d3aa3fb0504a9407f343e5613fd0512e", "messageHeadline": "[ROCm] Add ROCm Debug wheels to private ADO Feeds (#17887)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 5, "deletions": 6}}, {"node": {"oid": "07317316ccbbbacc4e8aa6fd50ba8563a482e226", "messageHeadline": "CUDA EP vs ROCM EP hipify audit (#17776)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 1512, "deletions": 1479}}, {"node": {"oid": "ba7f20ac573cc811b59b3aa621912ecf71b0d7a9", "messageHeadline": "Fix illegal opcode error from mlas (#17885)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "ca8cab29cd284ba65d1955854c7dcc76c0debe89", "messageHeadline": "distributed slice (#17761)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 449, "deletions": 29}}, {"node": {"oid": "3f3ece4a39a148c6d7ac29852ea3c43cb6264a08", "messageHeadline": "Update NDK to 26.0.10792818 (#17852)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "163218d6d791b014ebd62e65678f6ab50a1a5f53", "messageHeadline": "[WebNN EP] Update Op Softmax for readability (#17826)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 4, "deletions": 6}}, {"node": {"oid": "c373a808a2b31377b70025df3daa9dc43ea7e8e5", "messageHeadline": "Add \"glue\" between training WASM artifacts and training web (#17474)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 77, "deletions": 12}}, {"node": {"oid": "809c8905fe64809b3d43563652286d1d05ba109f", "messageHeadline": "Update TestCase.cc: exclude a test for DML (#17909)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "2efab54f9c7ee389dfa9a5f494b043d79290afd6", "messageHeadline": "Fix missing registration for new dml device selection API (#17898)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "fa0a79a921e60f3db2df3cc07f2e8ef858818acf", "messageHeadline": "Fix Triton Compile Error for Codegened Dropout Code (#17899)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9d07ca362136e2cc3a9d05f41de0d6bd197d989d", "messageHeadline": "Move compliance check before publishing pipeline artifact (#17857)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "74a8acf4057902f15b574046b2c6b0e3a0b11e49", "messageHeadline": "Set default value for NVCC threads (#17866)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "e2cd6748fc460bd9d5ee12d9d40193a0a1ab6b77", "messageHeadline": "Fix GroupNorm fusion: skip if num of channels not supported (#17869)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 13, "deletions": 8}}, {"node": {"oid": "25bbd8d4ebed60c28f702a9fe6b0e6a12c19de9c", "messageHeadline": "[js/web] allow gpu IO binding tests to fail temporarily (#17892)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "138ccecd22a2f85eff28e58759c06b33d7387e41", "messageHeadline": "Change how \"NPM packaging pipeline\" downloads packages from another p\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 31}}, {"node": {"oid": "20798a9f03c37a3d090d7d627c5986e1cf5ea568", "messageHeadline": "Enable onnx_test_runner to run the whole models dir in CI machine (#1\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 73, "deletions": 16}}, {"node": {"oid": "b3cab55d68cd887c1a43d44f951c93189cf91af6", "messageHeadline": "[WebNN EP] Add a duplicate entry to support new \"dataType\" (#17841)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 45, "deletions": 44}}, {"node": {"oid": "565bead85fccb978a40dbf8ec9908e022f5bd678", "messageHeadline": "[QNN EP] Support Softmax/LogSoftmax with any axis attribute (#17877)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 298, "deletions": 44}}, {"node": {"oid": "63dc5dc1a913c66644970717a9bc75b432d87954", "messageHeadline": "Add document for PythonOp (#17888)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 156, "deletions": 0}}, {"node": {"oid": "d532645beddb8a544ad70f90eb1ab26a473224f7", "messageHeadline": "[js/webgpu] revise uniform support (#17871)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 77, "deletions": 131}}, {"node": {"oid": "b8f373b0aee086a36ea357bc2ffb2944246be15a", "messageHeadline": "Add API for NPU Device Selection in the DML EP (#17612)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 390, "deletions": 27}}, {"node": {"oid": "a441a71e8e1b9f8cd1fa83cbc4d6aa5f4bec8e2f", "messageHeadline": "[js/web] support different export format for ort-web (#17878)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 184, "deletions": 44}}, {"node": {"oid": "0e2782438a65b97919f15af14d2a4ada361157b6", "messageHeadline": "Support inplace update for PythonOp/Grad (#17687)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 590, "deletions": 78}}, {"node": {"oid": "54b7503c3024682a58ebcdb5d823923ca420eba5", "messageHeadline": "create patch for allgather fn for deepspeed stage 3 (#17855)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 53, "deletions": 1}}, {"node": {"oid": "948c8369a0e3c5b82ef9e887df09b6fe92ecc7fb", "messageHeadline": "[CUDA/ROCm] Remove limitation of BiasAdd (#17848)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 30, "deletions": 10}}, {"node": {"oid": "5228332c9fe1303170f9c5976cf566eeb70c9b0e", "messageHeadline": "[js] upgrade JS shared dev dependencies (#17831)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1045, "deletions": 4590}}, {"node": {"oid": "c6f1a1ce696dbc85bc45d3f7b6bed6821304e282", "messageHeadline": "update build_jsep.bat to add release build flags (#17471)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d637111e9f3f816a81a8bfc27d7f183b824fc6a4", "messageHeadline": "[CUDA/ROCm] Update BiasSplitGelu for SD XL Refiner model (#17849)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 30, "deletions": 2}}, {"node": {"oid": "9a1c884ba3103b89d93183386b09e486cf71ea4a", "messageHeadline": "[QNN EP] Add script to generate Onnx model from native QNN generated \u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 151, "deletions": 1}}, {"node": {"oid": "d9b9c5a537bc51595eb30e2d4fff288ea7c2647f", "messageHeadline": "[js/webgpu] support using uniform buffer (#17803)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1213, "deletions": 1205}}, {"node": {"oid": "53be802f39687b29b32fbd5c842e18c7567a0a9e", "messageHeadline": "Onnx_test_runner and onnxruntime_test_all use the same broken test li\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 583, "deletions": 996}}, {"node": {"oid": "05ac9f6f2a9ccffe5ae6d65ddf56a0df06ac991d", "messageHeadline": "Split onnxruntime_providers.cmake to multiple (#17853)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1806, "deletions": 1724}}, {"node": {"oid": "046939b0c12c092beb84b8dd8ab2a7905f069f44", "messageHeadline": "Include CoreML in mac os python packages (#17844)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "9c716f4557ea48576efc466dd04000eb80155b6a", "messageHeadline": "Add noexcep_operators to onnxruntime internal libraries (#17850)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "406cd324e0f7de5ecb58e55c267924b790a49e85", "messageHeadline": "[CUDA] GroupQueryAttention operator using FlashAttention (#17674)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 3448, "deletions": 191}}, {"node": {"oid": "ba72bb6f98192341e56a6fe0b509d90f2b7e9f57", "messageHeadline": "Fix a typo in ABI_Dev_Notes.md (#17832)", "author": {"name": "kyoshisuki", "email": "143475866+kyoshisuki@users.noreply.github.com", "user": {"login": "kyoshisuki"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "60f19ab001b40d8e0bd2912645afba28d2c016ba", "messageHeadline": "Fix Pad's quantization (#17807)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 131, "deletions": 2}}, {"node": {"oid": "2ef6ee674cde2b7b61a5f50a1017db393b894269", "messageHeadline": "[ROCm] Update ROCm and MIGraphX CI to ROCm5.7 (#17834)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 13, "deletions": 49}}, {"node": {"oid": "c2bd5b70b29eb3c687c5497696e7b0a1930604d3", "messageHeadline": "Fix enable_training and use_migraphx (#17827)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "faf9a0f6c75f4f1e2fabdd67b5b6670e0d0cda26", "messageHeadline": "Fix runtime installation error (#17828)", "author": {"name": "MistEO", "email": "mistereo@hotmail.com", "user": {"login": "MistEO"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "b5a103ae1614c2cd9b96a15bb1ab1887d5eedd36", "messageHeadline": "Upgrade transformers to fix CI (#17823)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b76994dc3a98715f624a9263bb0ff123b03b1e25", "messageHeadline": "Improve CUDA EP's GetCapability (#17809)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 18, "deletions": 8}}, {"node": {"oid": "37f4f27da0dae38b225ccf0f9a2ab8b9290222c1", "messageHeadline": "[ROCm] ONNX Runtime training rocm package for ADO (#17683)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 24, "deletions": 11}}, {"node": {"oid": "7201def4ec0b6850c4ab243e0f210eece6b1b707", "messageHeadline": "Fix convergence for dolly+stage3 training (#17685)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 657, "deletions": 155}}, {"node": {"oid": "891b50cc683bf3437c04a228a17b4450fefbd9cf", "messageHeadline": "General INFO logging tracking occurance of GraphTransformer modificat\u2026", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "385fab5baec0ba0e936a7232cc5657bf85866af8", "messageHeadline": "[QNN EP] Qnn cache improvement (#17757)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 769, "deletions": 284}}, {"node": {"oid": "569876fb16cce5165035cca755c9e5ca2c86dcc6", "messageHeadline": "[TensorRT EP] Refactor OrtTensorRTProviderOptions initialization and \u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 445, "deletions": 485}}, {"node": {"oid": "6ea493571e2c0e9e507fc6c6c1d71a98e96011f9", "messageHeadline": "[js/web] use esbuild to accelerate bundle build (#17745)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1098, "deletions": 593}}, {"node": {"oid": "be1e51af2aa458c08dd41d0e4172fd3d3852e821", "messageHeadline": "Add length checks to fusion_transpose.py (#17608)", "author": {"name": "Kaz Nishimura", "email": "kazssym@linuxfront.com", "user": {"login": "kazssym"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "735df7e2a885141c654fe0c10ab5055e4ab22adc", "messageHeadline": "[webgpu]: add a simple GetCapability implementation (#17643)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 42, "deletions": 1}}, {"node": {"oid": "cb9408e89cdaad084f3ee614a928ad9a25948420", "messageHeadline": "Enable cpp20 builds for DML EP and WinML API (#17800)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 119, "deletions": 70}}, {"node": {"oid": "3878011ce2ba0cf15839cc9ddd9587904292334a", "messageHeadline": "Remove MPI dependency (#17624)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 191, "deletions": 33}}, {"node": {"oid": "b306b02a8689bc84237a82df96dd97a0b903d779", "messageHeadline": "[QNN EP] fixed input for InstanceNormU8 unit test and update copy lib\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 11, "deletions": 6}}, {"node": {"oid": "be7541ef4a39fe41a7a616c621c1fdd163c27035", "messageHeadline": "[Linter] Bump ruff and remove pylint (#17797)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 108, "deletions": 139}}, {"node": {"oid": "7417fd41e2c77cac5f09c3224918c9c23598a1cd", "messageHeadline": "[QNN EP] Add better unit tests for rank 5 ReduceSum (#17802)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 34, "deletions": 11}}, {"node": {"oid": "5be79e2e295ab6270f61f5658850dbdef1c9ebd3", "messageHeadline": "Remove swift files on ORT main repo (#17799)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 0, "deletions": 218}}, {"node": {"oid": "faef9c32fa7740935ee6cde5b00adab59d641854", "messageHeadline": "ONNX-Native Tensor Parallel: Using Distributed MatMul as Example (#17\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1969, "deletions": 11}}, {"node": {"oid": "742069a8e860b6056de051e13eb0e2d4b07dbd2a", "messageHeadline": "Add option for max intermediate outputs for MinMaxCalibrater (#17029)", "author": {"name": "Benedikt Hilmes", "email": "benedikt.hilmes@rwth-aachen.de", "user": {"login": "Atticus1806"}}, "additions": 20, "deletions": 2}}, {"node": {"oid": "b6bef0f06353f9b0fafd8a6c800dd0b336a9e179", "messageHeadline": "Add test for iOS dynamic framework (#17790)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 36, "deletions": 0}}, {"node": {"oid": "0e988239cc84d87c33fbea40cc2e80fa56ce5884", "messageHeadline": "[BeamSearch]optimize key cache reordering (#17771)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 85, "deletions": 23}}, {"node": {"oid": "e1a089c23c5f7c89d22a88464df81ade5342530a", "messageHeadline": "[QNN EP] Skip Op validation for Q & DQ node with 5D data (#17792)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "d6dad96923423d5a8ae184f3e88fdf48aa6aab57", "messageHeadline": "Add CUDA EP in StableDiffusion demo (#17788)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 341, "deletions": 53}}, {"node": {"oid": "db3901ab9790fef952aa6a0147eedf671c17c899", "messageHeadline": "[js/webgpu] Enable the NCHW ConvMatMul path (#17717)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 99, "deletions": 49}}, {"node": {"oid": "1bc115719c0d07ad658a9bb458ddf8f1bf379a41", "messageHeadline": "Unify handling of public headers in onnxruntime.cmake. (#17779)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 21, "deletions": 36}}, {"node": {"oid": "a05580ed5be3f6311c6184136f953eeb14f48dcd", "messageHeadline": "StableDiffusion XL with TensorRT EP (#17748)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 5090, "deletions": 1292}}, {"node": {"oid": "8e6019af2ebcca0bbb1da425bc882e74a3843015", "messageHeadline": "[QNN EP] Enable QNN Saver for debugging issues (#17747)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 259, "deletions": 81}}, {"node": {"oid": "992f3e460945070f7239d60ba4d5eff0b1ab5897", "messageHeadline": "[js/webgpu] Support where (#17544)", "author": {"name": "Xu Xing", "email": "xing.xu@intel.com", "user": {"login": "axinging"}}, "additions": 455, "deletions": 1}}, {"node": {"oid": "f8a8452a6b4d494c9d60eda6708309bcbcdea1fa", "messageHeadline": "[js/webgpu] fix pad operator (#17775)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d0519a760326fc7d87afdcbc1ea4c110ecbe3ce8", "messageHeadline": "[js/web] BiasSplitGelu and BiasAdd kernels (#17161)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 2443, "deletions": 0}}, {"node": {"oid": "d11e053412960fc677d45215a0b3f8d9e23fc53c", "messageHeadline": "Add option to specify the EP to use, enabling DML EP and others (#17490)", "author": {"name": "Kaz Nishimura", "email": "kazssym@linuxfront.com", "user": {"login": "kazssym"}}, "additions": 54, "deletions": 8}}, {"node": {"oid": "451c02543a3a696d5ebe5ad5ef38f9298e6c1bb0", "messageHeadline": "[js/webgpu] allow specify preferredLayout (#17756)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 40, "deletions": 5}}, {"node": {"oid": "f158f394d695a196ae2e06b350523a7dc741321d", "messageHeadline": "[WebNN EP] Support Softmax since version 13 (#17714)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 69, "deletions": 28}}, {"node": {"oid": "ac4e72604605be524f70d1e92da81b70c5db984a", "messageHeadline": "Add bytes model loading test to react native e2e (#17749)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 16, "deletions": 3}}, {"node": {"oid": "63acaf47d2193cbe175f5cabf0cb20b25d6e39aa", "messageHeadline": "Fix onnx quantizer activation and weight type attribute (#17651)", "author": {"name": "Ella Charlaix", "email": "80481427+echarlaix@users.noreply.github.com", "user": {"login": "echarlaix"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "0d606046380caa32fddffb0ee4d3434418719ede", "messageHeadline": "[JS/WebGPU] support Range operator (#17233)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 130, "deletions": 5}}, {"node": {"oid": "a941dd583e6061d57564d770ae727860ff9b237e", "messageHeadline": "[js/web] FP16 Conv, ConvTranspose and MatMul (#17514)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 148, "deletions": 174}}, {"node": {"oid": "9aad78721c3873e6c9b148584089d3a13a889fc6", "messageHeadline": "Update debug_alloc.cc: filter out one more memory leak from absl (#17\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "668c70ee11b6b20c56997a9bc68e93317674e803", "messageHeadline": "Add support for specifying a custom logging function per session. (#1\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 160, "deletions": 49}}, {"node": {"oid": "6a5f469d44aca607bd08cc2aca117c33bab31da8", "messageHeadline": "Add training interfaces to js/common (#17333)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 243, "deletions": 29}}, {"node": {"oid": "e106b1eb8f22c57414a8bb4e69cbb62b702b12c7", "messageHeadline": "Fix react native load from Uint8Array buffer bug (#17739)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "5a623dca0118a1fc75419875d2af813161d097a0", "messageHeadline": "Python API to check whether collective ops are available or not (#17730)", "author": {"name": "shaahji", "email": "96227573+shaahji@users.noreply.github.com", "user": {"login": "shaahji"}}, "additions": 22, "deletions": 0}}, {"node": {"oid": "14d349e29075db52eb6971569b572ebe54215596", "messageHeadline": "Enable backtrace in unit tests (#17655)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 83, "deletions": 85}}, {"node": {"oid": "561aca97cfcf76ce6d190a2403cae34c17bee75a", "messageHeadline": "[js/webgpu] support IO binding (#17480)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1177, "deletions": 288}}, {"node": {"oid": "b4fbc25b1f5c3942b59ae2a75a72d070f92d7bb1", "messageHeadline": "[JS/Web] Add ConvTranspose implementation using MatMul (#17573)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 471, "deletions": 19}}, {"node": {"oid": "caf98128c1dbbee1ce3c831364cee88031471a32", "messageHeadline": "Update linux-wasm-ci.yml: remove the ln command (#17735)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "9cb60c5b86cd13aab1b782915e1ac1f794fd7fe1", "messageHeadline": "Resize and EP specific transpose optimization updates (#17664)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 235, "deletions": 115}}, {"node": {"oid": "20f96fd096e8d3b866c0e947d420f971234eb795", "messageHeadline": "Fix Attention Runtime Error for CLIP model (#17729)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 22, "deletions": 20}}, {"node": {"oid": "fc9a69dcae4f81839593feaa69d081edf9b3c564", "messageHeadline": "Update VecAddMoveOnlyFunctor and VecAddWithIsSupportedMethod with Def\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "91367484627a25b46520674dcc320efe505437ef", "messageHeadline": "Fix: Fail to skip disabledmodel in winml (#17728)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "1f4a3529ddde7796cd018b6c2bd97a6e099c4b88", "messageHeadline": "Bugfix: Add initializer to model in AttentionMask directly (#17719)", "author": {"name": "Jambay Kinley", "email": "jambaykinley@microsoft.com", "user": {"login": "jambayk"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "8621bdcd444f99891eaf8ab4ea7f5f60431400ba", "messageHeadline": "Remove unnecessary #incldue <stacktrace> (#17716)", "author": {"name": "MistEO", "email": "mistereo@hotmail.com", "user": {"login": "MistEO"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "870b0bc305e163801b47dc4989927a2274cf1e07", "messageHeadline": "Fix typo of cmake (#17715)", "author": {"name": "MistEO", "email": "mistereo@hotmail.com", "user": {"login": "MistEO"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "522cc968e8b05bb3c4b8dd4e87d1303743704eef", "messageHeadline": "[java] Filling out the javadoc for the float8 types (#17694)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 31, "deletions": 8}}, {"node": {"oid": "13b0f8a6ce448f2f6fa59a243905219cbd952ec4", "messageHeadline": "fix: supported typo (#17216)", "author": {"name": "Mustafa Ate\u015f Uzun", "email": "mustafauzun0@gmail.com", "user": {"login": "0o001"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "276e8733bda76e6e42845af1b795fcb8189aba96", "messageHeadline": "Update onnx python package and setuptools (#17709)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 33, "deletions": 22}}, {"node": {"oid": "e6aa0fa17464e786b5e0a9811de7a04fe86329bf", "messageHeadline": "Add Gelu Related Ops to Triton Codegen (#17713)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 110, "deletions": 20}}, {"node": {"oid": "a99c965d05007ecd9fb5cf5508c3200985690cb7", "messageHeadline": "Make transpose optimizer able to look past DQ node for const initiali\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 607, "deletions": 93}}, {"node": {"oid": "33295ed883d2a53c6e4f0a58c95729696feb3d5e", "messageHeadline": "Handle string initializers in constant folding (#17422)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 341, "deletions": 229}}, {"node": {"oid": "bcc6205161969b292c66601ef9b080dd418d21a2", "messageHeadline": "\ud83d\udc1b Bugfix win del file err (#17697)", "author": {"name": "trajep", "email": "trajepl@gmail.com", "user": {"login": "trajepl"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "2be4dc6d041e53ff926e5fdc5662b725ced126fa", "messageHeadline": "ONNX 1.15 integration (#17125)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 278, "deletions": 245}}, {"node": {"oid": "37dcefb5b7ed620b823f6638c178eac8b6e0ed9a", "messageHeadline": "Patch lite custom op API (#17605)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 656, "deletions": 211}}, {"node": {"oid": "4ab0e17fe812f712dafbb9e16d9d418365024d1e", "messageHeadline": "[Technical docs] Fixed a couple of old links in `FAQ.md` (#17415)", "author": {"name": "Nicol\u00f2 Lucchesi", "email": "nicolo.lucchesi@gmail.com", "user": {"login": "NickLucche"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "93f22aa189befb54ca6e8297387815dd8198f1f9", "messageHeadline": "[WebNN EP] Fix bug in Softmax (#17665)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 16, "deletions": 10}}, {"node": {"oid": "614af3742c3f1f3f2aa91a5699a6de1c9fe79fbb", "messageHeadline": "Add prepacked weights container to subgraphs (#17671)", "author": {"name": "Brian Lambert", "email": "98757707+brian-pieces@users.noreply.github.com", "user": {"login": "brian-pieces"}}, "additions": 126, "deletions": 1}}, {"node": {"oid": "0141e27ca1e9954da2630f817709e2eebcd88dde", "messageHeadline": "Enabling c++ 20 in MacOS build (#16187)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 96, "deletions": 87}}, {"node": {"oid": "b8e348145cbcfaa72172aa78b8c21f1544d3854c", "messageHeadline": "fixed #16873 (#16932)", "author": {"name": "Vadym Stupakov", "email": "vadim.stupakov@gmail.com", "user": {"login": "Red-Eyed"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "f43acf2d33ca4c2f87b0927929123ebfaed82b1a", "messageHeadline": "Close the JSON object in settings.json (#17583)", "author": {"name": "Kaz Nishimura", "email": "kazssym@linuxfront.com", "user": {"login": "kazssym"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "1c245e6775201fec7a1d269a71deb0f3af6f7bea", "messageHeadline": "Stop throwing exception on python binding when multiple EP available \u2026", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 9, "deletions": 33}}, {"node": {"oid": "7572e6055ce065cb135f3d4dd0bafa8f589bc7df", "messageHeadline": "[TensorRT EP] Back out the PerThreadContext  (#17690)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 55, "deletions": 82}}, {"node": {"oid": "aed43f429a961e2fee543b801da50899a11b43cb", "messageHeadline": "[java] Enable output pinning in OrtSession and OrtTrainingSession (#1\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 668, "deletions": 119}}, {"node": {"oid": "ccb73fd827d29f69b1b5dfcbd4c27188b2364f0d", "messageHeadline": "[On-Device Training] Expose Parameters through the Training API (#17364)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 936, "deletions": 159}}, {"node": {"oid": "95e8dfaea51f6e6dda89b66d95c672927f578f21", "messageHeadline": "Update quant_utils.py/write_calibration_table (#17314)", "author": {"name": "aimilefth", "email": "60664743+aimilefth@users.noreply.github.com", "user": {"login": "aimilefth"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "a942bbf4897688fd1bd88bb1db73d2f8d648bcbc", "messageHeadline": "Update nodejs to 18.x (#17657)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 117}}, {"node": {"oid": "b2b140860817d974b5f9dd0b8943c49a725a5edd", "messageHeadline": "[js/web] update browser launch cmd flags (#17658)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 9}}, {"node": {"oid": "f50fa46fe09819734550a7e8725db999210e2b97", "messageHeadline": "[JSEP] allow DataTransfer to deal with zero sized input (#17661)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 18, "deletions": 16}}, {"node": {"oid": "fcfc2391b818af5e28ded7f669a2f466f9276cb1", "messageHeadline": "[JSEP] allow JsCustomAllocator to deal with zero sized input (#17660)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "905faea3b2683383bdb71f4ef3bb0a8f0a0832c2", "messageHeadline": "Fix static quantization for QDQ and Percentile distribution (#17649)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 13909, "deletions": 7}}, {"node": {"oid": "df15a3a335cb4e6703404beeb405f987521e4cf9", "messageHeadline": "[js/web] configure 5GB memory space for webpack build (#17684)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "216214b7d302cb504d1e5a647f65b6fe49c22dbb", "messageHeadline": "[ROCm] Remove ROCm5.4.2, ROCm 5.5 and add ROCm5.7 to python package p\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 9, "deletions": 22}}, {"node": {"oid": "ce287a4e77895e7f6147a044ae5c723a48cb8277", "messageHeadline": "[WebNN EP] Remove workaround for dynamic shape (#17644)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "e70a23f8dc6fc181218106f0e12730f980cc867e", "messageHeadline": "[QNN EP] Integrate Resize op fixes from QNN 2.14.1 (#17641)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 308, "deletions": 333}}, {"node": {"oid": "6d7bc2a097a1a08541cd0d4628831c79ab8092d5", "messageHeadline": "Fix ARMv7 build (#13891)", "author": {"name": "Lukas Berbuer", "email": "36054362+lukasberbuer@users.noreply.github.com", "user": {"login": "lukasberbuer"}}, "additions": 25, "deletions": 29}}, {"node": {"oid": "55b16d347cbcde41b35c3ed12f34eeca1a1b05d6", "messageHeadline": "Read model zoo test (#17666)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "891fba3b9cd71e2e1afdeab9fb3c5b5497db20cf", "messageHeadline": "[js/webgpu] Optimize Gather op (#17625)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 42, "deletions": 51}}, {"node": {"oid": "cd3fb377ea867570796cf61bc420cd985129a2a0", "messageHeadline": "[js/webgpu] Allow binary ops with scalar to use the vectorize path (#\u2026", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 17, "deletions": 6}}, {"node": {"oid": "1bc215e1d1c1e3509a1dd0bc413b1537563dedb5", "messageHeadline": "[VITISAI] add float16 and bfloat16 support (#17438)", "author": {"name": "Yiming Hu", "email": "woinck@users.noreply.github.com", "user": {"login": "woinck"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "6b7bce5ec992f2b3333ee22066201f53e7978faf", "messageHeadline": "Model post process for zero stage3 training (#17187)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 619, "deletions": 171}}, {"node": {"oid": "498b60d8a47b398fefeb1847cadde07bef6e99fa", "messageHeadline": "[js/web] fp16 Pool & Reduce (#17512)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 206, "deletions": 220}}, {"node": {"oid": "d56fc7ebf5377abc96db728eafaffd8bf79a3b81", "messageHeadline": "Layer norm fusion deepspeed stage3 changes (#17614)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 136, "deletions": 21}}, {"node": {"oid": "f299016cbe87a5341e0a8aa69b621555c9d49a35", "messageHeadline": "Fix crash on Windows server 2016 on Intel Gen4 Xeon processors (#17611)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 11, "deletions": 6}}, {"node": {"oid": "5b9cd91a9cddbe7c461c1ad7ca44edd5111ea920", "messageHeadline": "[ROCm] fix CI (#17648)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "57dfd15d7bc9d9c5779896f6685ec473875dc6e1", "messageHeadline": "Remove dnf update from docker build scripts (#17551)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 351, "deletions": 244}}, {"node": {"oid": "038c76378fdee45261d43af45466a0797e6ad124", "messageHeadline": "Include onnxruntime_float16.h in the package. (#17637)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "4f3f4366d5838f80b45060c26bcd648ee387a252", "messageHeadline": "Fix API 16's marker (#17640)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "1f991f27f1d4a8d19c877bc4d33457dd45994c80", "messageHeadline": "[ROCm] add manylinux build test for ROCm CI (#17621)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 111, "deletions": 3}}, {"node": {"oid": "dd561f201524ea5c78d9fa26df818397139f80af", "messageHeadline": "Upgrade sympy (#17639)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "c55da45e20435b8aa9edb78179b6027502b778b0", "messageHeadline": "[QNN EP] Add more op unit tests (fix Clip, TopK, Tile) (#17457)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 2273, "deletions": 316}}, {"node": {"oid": "c65e892089e2e6f383e80339334450ac06be5ba0", "messageHeadline": "[CUDA] Fix performance bug in DecoderMaskedMultiheadAttention for Bea\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 8, "deletions": 11}}, {"node": {"oid": "e6301eee6aa3f22f31bdcf431d0ed0ca828d20be", "messageHeadline": "Bump Up Version to 1.17.0 (#17587)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 31, "deletions": 26}}, {"node": {"oid": "f297d4dfb9f181942a224c96186d07db7155e61e", "messageHeadline": "Remove onnxruntime extensions from list of gitmodules (#17615)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "d522cc7cc47e6651446a1104d8f94e3b37b34ea6", "messageHeadline": "Update npm-packaging-pipeline.yml to always use artifacts from main b\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "460f17fbb84e29cd693c6009199070e0851aae0d", "messageHeadline": "[JS/WebGPU] Support If on WebGPU (#17478)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 108, "deletions": 4}}, {"node": {"oid": "152e61da37514e0b43bf1d9929ac6b43d54859be", "messageHeadline": "Avoid `get_logger` overriding root logger level (#17569)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "730fab3050c631b6a888990f55840ed6481bd07b", "messageHeadline": "Refactor Attention cuda kernel (#17578)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 904, "deletions": 728}}, {"node": {"oid": "068300d97eb25e5b52324e7af54a45ed1fa6a4c3", "messageHeadline": "Pin beartype version (#17599)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "d350ab31d7ca5a6649022ef426caf51e56430bfe", "messageHeadline": "Remove reference to internals in torch.onnx in test (#17550)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "f969e7f8d866585e79c72df8a9c043f7e5be69bb", "messageHeadline": "Provide kwargs to remove_shared_initializers (#17539)", "author": {"name": "Jambay Kinley", "email": "jambaykinley@microsoft.com", "user": {"login": "jambayk"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "7116e66c4b575ed816b2c836132e183b594c173a", "messageHeadline": "Improve Win QNNEP pipeline (#17586)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 41, "deletions": 21}}, {"node": {"oid": "0f406ca1d364d6a85d7a82f92f83bfecc04d94bd", "messageHeadline": "[js/web] FP16 binary and unary ops (#17515)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 81, "deletions": 69}}, {"node": {"oid": "dea425e7c140a7216727421c434a1c5a3fb1f54d", "messageHeadline": "[QNN/CPU EP] Add 16-bit Quantize/Dequantize contrib ops (#17015)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 2837, "deletions": 1211}}, {"node": {"oid": "af14ae8050aa77d40031e7794fb2bf38bf30acef", "messageHeadline": "[ROCm] Update whisper benchmark script (#17391)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 198, "deletions": 91}}, {"node": {"oid": "c96923732141bf3823f1b5acb666c407f07fa9c2", "messageHeadline": "[TRT EP] Fix ProviderOptions functions (#17567)", "author": {"name": "simonjub", "email": "78098752+simonjub@users.noreply.github.com", "user": {"login": "simonjub"}}, "additions": 22, "deletions": 3}}, {"node": {"oid": "705f8a371886828dcd2a380d10ba3a6549a60b9b", "messageHeadline": "[TensorRT EP] Fallback to CUDA EP if it's explicitly assigned (#17535)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 20, "deletions": 3}}, {"node": {"oid": "efd416b71f0925108ff7cb0e83e99beabe7c05cd", "messageHeadline": "[js/web] update test to explicitly fail for webnn without proxy (#17554)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "155887593db4e95d3b808ec605e9b089daab4ef0", "messageHeadline": "[js/web] update npm test to load test cases only for required backend\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "fdb132643d8e07d383d6cd9af1c19f547fe1718f", "messageHeadline": "Remove redundant Resolve() after each inlined function (#17556)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "adb0be45d38e83609e9941917d233bb43404ff36", "messageHeadline": "Refactoring of attention cuda kernel: move prepare qkv and concat_pas\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 729, "deletions": 674}}, {"node": {"oid": "af80542e65334979d3e37c1cb6c88068c5c65f3b", "messageHeadline": "Update optimize_pipeline for SDXL (#17536)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 226, "deletions": 69}}, {"node": {"oid": "377f959c69e9f213cd4a8c71a5e80162a412989a", "messageHeadline": "Run Final_Jar_Testing_Linux_GPU in docker (#17533)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 82, "deletions": 12}}, {"node": {"oid": "a5302fec93cd91d976e3798e38873927a2ddd8a9", "messageHeadline": "[WebNN EP] Fix bug for PRelu on CPU backend. (#17543)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 22, "deletions": 0}}, {"node": {"oid": "4d931edd78370b18805bd908e92c28e50994f6af", "messageHeadline": "Update tensorrt_dependencies in setup.py (#17562)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "94f2ed6bbd95d0561eaad7c7a42ee3f56eb9f29e", "messageHeadline": "run_CIs_for_external_pr.py: update required pipelines (#17557)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "9aafbe3febd569da2613230ce45b1ce1054488d5", "messageHeadline": "[js/web] revise TensorView (#17473)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 72, "deletions": 149}}, {"node": {"oid": "a2fba28f6cdc6a4ea946a9d56b4b4d19241ebe27", "messageHeadline": "Remove extraneous javascript includes (#17558)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3a1e48dd5ad2a9ef95b6bc137441a4df6b05e5d0", "messageHeadline": "update BERT notebook with ORT 1.16 (#17524)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 956, "deletions": 477}}, {"node": {"oid": "5ed5f13920b66098cf3bfe7da85f463e6ff64bac", "messageHeadline": "[DML EP] Add missing member initializer DmlGraphNodeCreateInfo::nodeC\u2026", "author": {"name": "Kaz Nishimura", "email": "kazssym@linuxfront.com", "user": {"login": "kazssym"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "41d2ff622c49aa3628c04f6b64ed7f33c8d80f30", "messageHeadline": "[js/webgpu] Optimize InstanceNormalization (#17491)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 147, "deletions": 53}}, {"node": {"oid": "46fe08226feb95ce03b6a4773d59b9786f14a3df", "messageHeadline": "[QNN EP] Enable Pad op support for QNN EP (#17508)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 651, "deletions": 4}}, {"node": {"oid": "198d4688490e8dfd447ba2cc7c99bb30c9b0ef92", "messageHeadline": "[WebGPU/JS] Added Pad operator support (#16928)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 379, "deletions": 5}}, {"node": {"oid": "e11849e716d7d63d0ff2eee8e66b93406c1ef547", "messageHeadline": "Configure StringNormalizer `default_locale` for _APPLE_ system  (#17339)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "7af2f68ef307a66a770a620ad2611e511bb2008f", "messageHeadline": "[js/web] add a test flag to customize chromium flags (#17545)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 62, "deletions": 57}}, {"node": {"oid": "5af6279440a0db698b0afbfc3de655400562831f", "messageHeadline": "Fix Android build (#17540)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 45, "deletions": 3}}, {"node": {"oid": "ad369a1fadb9939b615c442f2a9f40ef5e7b6ec0", "messageHeadline": "[js/rn] Support create boolean tensor (#17052)", "author": {"name": "Hans", "email": "me@hans00.me", "user": {"login": "hans00"}}, "additions": 33, "deletions": 1}}, {"node": {"oid": "32f5658abb5f2fc82c981b623d8e7969eb46e14f", "messageHeadline": "remove gsl to make status.h independent from gsl (#17402)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "03b56f7a73319e95e381a623325207665c6ac037", "messageHeadline": "[js/webgpu] FP16 extension registration (#17493)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 17, "deletions": 7}}, {"node": {"oid": "7edff1c2bfca812ce352b8df8efc6c904c4292ca", "messageHeadline": "[DML EP] Add subgraph fusion support (#17504)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 76, "deletions": 33}}, {"node": {"oid": "4e37c5d1f0c54d43ba2662f490d84cd0333c8813", "messageHeadline": "Bump actions/checkout from 3 to 4 (#17487)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 15, "deletions": 15}}, {"node": {"oid": "a2e75114cc9c38ab4c48fac4c884d53f2d0d17d4", "messageHeadline": "[js/web] add sessionOptions.freeDimensionOverrides (#17488)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 36, "deletions": 0}}, {"node": {"oid": "5d3786206bbe8a84b9d3f9c62e310258e3570ff2", "messageHeadline": "Fix ROCM's nightly build (#17518)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "54a092c42714caa0d236e0f67b07b0804dbb9490", "messageHeadline": "[DML EP] Complete python IO binding implementation (#17344)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 413, "deletions": 254}}, {"node": {"oid": "c0a4fe777fcc1311bf1379651ca68dfde176d94d", "messageHeadline": "Move Linux python test into docker  (#17479)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 73, "deletions": 71}}, {"node": {"oid": "cdf3e9dba9718b4b461540894905e01357096b7d", "messageHeadline": "[js] update prepack script to use exact version (#17484)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "ec94b07f0a14fc0a805e767803d483059cbd58ff", "messageHeadline": "[JS/WebGPU] support Concat.int32 operator (#17003)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 428, "deletions": 17}}, {"node": {"oid": "41584b28278a25f0e6a269a47264bccad9888f87", "messageHeadline": "[js/web] ensure ORT initialization to run only once (#17529)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "24a3c740c06f3c3b2d97f2839ece55bbd8ea0dd9", "messageHeadline": "Revert \"[ROCm][MIGraphX] for googletest dep, set OVERRIDE_FIND_PACKAG\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b52127d22d008254f20de1956e698b8a6f8712d6", "messageHeadline": "update acpt image for the training ci nightly (#17521)", "author": {"name": "rui-ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bbcf4b45dccb5ef7c2c8b516feb89505696ff602", "messageHeadline": "Upgrade doxygen to 1.9.8 (#17525)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 325, "deletions": 95}}, {"node": {"oid": "9b755dce9f6b39296a716028f52347de037daa2e", "messageHeadline": "Delete all Prefast tasks (#17522)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 318}}, {"node": {"oid": "f923eec28b6b4237d6fc251042893deab335d97e", "messageHeadline": "[js/web] release session after use in npm test (#17470)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "49511b548371b2af4d9c2d186204dd85dc0a1a7e", "messageHeadline": "Improve performance of prune_graph in onnx_model.py (#17502)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 52, "deletions": 26}}, {"node": {"oid": "cf672c5887b4e5991b022df38fbb8d61f29fe420", "messageHeadline": "Use name of temporary provisioning profile. (#17459)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "aa5e36456a5720fc341f59ca3fbd913c10f501e7", "messageHeadline": "[TRT EP] Fix multithreading bug of getting the corrupted trt engine i\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 200, "deletions": 202}}, {"node": {"oid": "db558ef9b47894179346472592d145448574197a", "messageHeadline": "TreeEnsemble speed up (#17449)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 194, "deletions": 165}}, {"node": {"oid": "65249f42e49c4ea98ccba9cf1340f76b896ef1f1", "messageHeadline": "[js/web] FP16 Gemm, Softmax & Transpose (#17494)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 73, "deletions": 88}}, {"node": {"oid": "f20e475e678706b1521fc412557b91763b4f5571", "messageHeadline": "[QNN EP] Update QNN SDK to version 2.14.1 (#17467)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 258, "deletions": 158}}, {"node": {"oid": "8ad9ab1b9ae716a9067055190b3ab1a5cd0995fa", "messageHeadline": "Fix CPU constant folding not reverting the node to its previous EP (#\u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 59, "deletions": 7}}, {"node": {"oid": "bf6d6961ccff555a0fbb6d208f81dc206e5ad11b", "messageHeadline": "[JS/Web] Added Einsum operator support. (#17401)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 979, "deletions": 3}}, {"node": {"oid": "850baced337365913a2fc8c3f88edc94014a1d45", "messageHeadline": "[web] a few updates to web pipeline (#17485)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 15, "deletions": 158}}, {"node": {"oid": "24f0893d3c81eb09b2e8c91793329d45adc27570", "messageHeadline": "Enable optimize_by_onnxruntime call for float32 unet model (#17483)", "author": {"name": "Kaz Nishimura", "email": "kazssym@linuxfront.com", "user": {"login": "kazssym"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "b827ab0efc2873031f6f6021461e5062e0f33434", "messageHeadline": "[TRT EP] Fix build error for building oss onnx-tensorrt parser (#17468)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "89da5a01080797a227ad55f004cc386230f91ecd", "messageHeadline": "[js/webgpu] exclude WebGPU reduce_log_sum_exp_* float64 test cases (#\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 9, "deletions": 8}}, {"node": {"oid": "550293d9adcabd0739f862d62c5f6e8d8da00c70", "messageHeadline": "OrtMemoryInfo: support new name \"WebGPU_Buffer\" (#17469)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 14}}, {"node": {"oid": "dcc93909b46302e213a9a4d9f9fde9231ddb70ee", "messageHeadline": "Add training WASM generation to Web CI pipeline (#17319)", "author": {"name": "Caroline Zhu", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 34, "deletions": 9}}, {"node": {"oid": "29a818caa09ad377d6cc019e11f90630e1eaaf66", "messageHeadline": "Attention fusion for stable diffusion clip model (#17445)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 403, "deletions": 15}}, {"node": {"oid": "4d753b74a5bc7101fe32c83fcc5690c8d226c2b6", "messageHeadline": "[js/common] prepare work for supporting webgpu IO binding implementat\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 40, "deletions": 33}}, {"node": {"oid": "bc84f52633444a2fef336f6eeb3113a8a5c19768", "messageHeadline": "Update C/C++ dependencies: abseil, date, nsync, googletest, wil, mp11\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 156, "deletions": 252}}, {"node": {"oid": "f51a765e6436e4c8090b40fb8281a22014f4a299", "messageHeadline": "Avoid calling patchelf (#17365)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 33, "deletions": 101}}, {"node": {"oid": "c5dbd5c91963481331dd4a91da1499c277585175", "messageHeadline": "Updates to training pipelines (#17292)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 21, "deletions": 112}}, {"node": {"oid": "ae74a517b62baa6d973e46b5b51ac9a640512c46", "messageHeadline": "Run Nuget_Test_Linux_GPU in container (#17452)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 121, "deletions": 23}}, {"node": {"oid": "7bc6dcecf7b5940de1eff2e93c0181440908d1fd", "messageHeadline": "fix embed layer norm fusion with embedding sum output (#17460)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 126, "deletions": 43}}, {"node": {"oid": "9017ea131b2215c094bfa042dd70bc2e5d543cb4", "messageHeadline": "[js/webgpu] support GreaterOrEqual and LessOrEqual operators (#17310)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 33, "deletions": 0}}, {"node": {"oid": "eaef4854617814659f53c502e483bb4e75a27f16", "messageHeadline": "Bump electron from 23.1.2 to 23.3.13 in /js/web (#17436)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "21c202bb5d57636f6f333620788a67d5359687cf", "messageHeadline": "Eliminate hashmap copies during function inlining (#17439)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 22, "deletions": 19}}, {"node": {"oid": "024f1dd72bf1a03babf66392d225d1d8d7d7fd2d", "messageHeadline": "Fix float 8 rounding on CPU (#16940)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 186, "deletions": 71}}, {"node": {"oid": "0a3eb60b017f2a7d691f0a3ce155f42a59d63b6c", "messageHeadline": "Fix Bug: Step failed but not exited with error  (#17442)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "b38fb0da0640df63b0978848e6b173da00e9ccc9", "messageHeadline": "Revert the yaml file changes in \"Nodejs_Packaging_CPU\" build job (#17\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 10}}, {"node": {"oid": "1e4bfa1da2a0885a0d08b61654f3cb75dad236be", "messageHeadline": "[QNN EP] Add more op unit tests (#17424)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 703, "deletions": 448}}, {"node": {"oid": "ede339f304c2d022d38ad2654ac7312fc71044f0", "messageHeadline": "Move dotnet build and test into docker in Linux CPU CI (#17417)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 43, "deletions": 44}}, {"node": {"oid": "7862a521b3bc18299a3f3154f0caf6c608f10117", "messageHeadline": "Update cmake's hash in android custom build (#17435)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e8b8d0d13b4394a10aae4c79976188e1c46bf426", "messageHeadline": "Fix weight tensors in transformers optimizer not saved to external da\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 161, "deletions": 159}}, {"node": {"oid": "2629cb8606c878caa48681ebd7559fb885c0f313", "messageHeadline": "[VitisAI] graph_save only saves proto of the graph instead of entire \u2026", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "8914fe687b3dab1fb46bec37c6a46bbbbefcb860", "messageHeadline": "[js/webgpu] Include Support for neg.int32 (#17374)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 30, "deletions": 3}}, {"node": {"oid": "a3a12372705c585240313d65916ab4a128caa4e3", "messageHeadline": "Disable xcpretty filtering of xcodebuild output in iOS packaging pipe\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "fa868ca9cd155468794068c053dc5c8190d1b9cd", "messageHeadline": "[js/node] release sessions after use in npm test (#17353)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "d88406a31bcd8fdf0b57b078f322d3cf6c62333d", "messageHeadline": "[js/common] use Map instead of object for backends (#17352)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "75710f0006b8105c8794f5017bb974c5deefd01d", "messageHeadline": "[js/webgpu] add matmul broadcast tests (#17335)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 220, "deletions": 0}}, {"node": {"oid": "110a2d0b73c9829f0a173a24284fe99ee8084830", "messageHeadline": "[build][wasm] add js_internal_api.js to link dependency (#17407)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "2cb75420ac6df4c61538c6ffd43817e17260ad35", "messageHeadline": "[js/common] clean up JSDoc (#17408)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "deda5db2315a1309b1f35c101c379c33de33eff6", "messageHeadline": "[ORTModule] Add Manual Seed to Fix UT Failure (#17411)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "16eba537a8f4ad7f32169661ffe2575cfb1f807a", "messageHeadline": "rust bindings: Do not unnecessarily re-run build.rs (#17018)", "author": {"name": "James Baker", "email": "j.baker@outlook.com", "user": {"login": "j-baker"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "c6b0d185b47e2b4ad48425a2302383915b2a9f03", "messageHeadline": "Update cmake to 3.27 and upgrade Linux CUDA docker files from CentOS7\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 381, "deletions": 244}}, {"node": {"oid": "026672e9472061f81abbcab0a712ea63e0e48acc", "messageHeadline": "[js/webgpu] Support slice int32 (#16968)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 49, "deletions": 4}}, {"node": {"oid": "e1a9f2ed6db83803757794aa4967486781c9dcae", "messageHeadline": "Fix insufficient space error in Android CI (#17423)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "fa28359beb88ec08d24090ae51ecac3c22fcc158", "messageHeadline": "Reduce GPU memory for Whisper models converted to ONNX (#17378)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 70, "deletions": 21}}, {"node": {"oid": "dbcc60bed5b094e0e769f6b4ef957918541000bb", "messageHeadline": "Introduce output type/shape validation (#17301)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 189, "deletions": 149}}, {"node": {"oid": "8818a99c93ec0be39e0b23be6aaf552e1dda9357", "messageHeadline": "Set proper nvcc threads to avoid OOM (#17419)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 45, "deletions": 11}}, {"node": {"oid": "e3bb2a0cdd81ec2527d12d4148bec02fcd947b95", "messageHeadline": "Fix git working dir for ORT_BUILD_INFO (fixes #17197) (#17198)", "author": {"name": "Lennart  Hannink", "email": "lehafb15@gmail.com", "user": {"login": "1enn0"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "6ea3908db482d4e62e16d9f2285928c0b9e48d2c", "messageHeadline": "Add ck's streamk and splitk gemm impl (#17280)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 97, "deletions": 3}}, {"node": {"oid": "5e747071be882efd6b54d7a7421042e68dcd6aff", "messageHeadline": "[js/webgpu] Fix bug in conv2dByMatMul path (#17369)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 39, "deletions": 39}}, {"node": {"oid": "e745575187acd64a1f7ce2b321e05ed6985363f6", "messageHeadline": "fix assert error in attention fusion script (#17375)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "e23f16adbfded27b88ac5a8b4f6f32beef05cb15", "messageHeadline": "output all parameters in the bert_perf_test tool (#17379)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "8b98ecad70c0f26d4cd43e41c5d96adad9cf2251", "messageHeadline": "Change RuntimeError to ImportError (#17380)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "16cfcd05907087679131e97447439704472773aa", "messageHeadline": "Fix NNAPI optional input handling checks and unblock Android CI pipel\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 70, "deletions": 28}}, {"node": {"oid": "44101e877125eaa18e191793973a4e1a002c6eca", "messageHeadline": "Flash Attention v2 MHA (#17227)", "author": {"name": "aciddelgado", "email": "139922440+aciddelgado@users.noreply.github.com", "user": {"login": "aciddelgado"}}, "additions": 4541, "deletions": 477}}, {"node": {"oid": "b54619509ff03e4c7d3b1b59c1531baef15150a4", "messageHeadline": "Refine build script for adding disable selected data types option (#1\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 19, "deletions": 17}}, {"node": {"oid": "30a450dcf8f90006b18a826199aefd0e7a6908b2", "messageHeadline": "Fix TRT EP's cuda graph feature (#17355)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 9, "deletions": 17}}, {"node": {"oid": "ae90b716ff60010c837f283be890b9c6613efabe", "messageHeadline": "Change _mm512_setzero to _mm512_setzero_ps (#17362)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "3a53836836fb1746a4f9e7d6b38ba0ee7c90028f", "messageHeadline": "[WebNN EP] Fix compilation with newer flatbuffers (#17367)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e60493525f60d4363c07551176e70daf782388fa", "messageHeadline": "[js/webgpu] Adding support for abs with int32 type (#17359)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 42, "deletions": 2}}, {"node": {"oid": "352b745debe88643092ed8787275335dab0a4571", "messageHeadline": "[js/webgpu] Add input/output shapes information to profiling (#17342)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 15, "deletions": 4}}, {"node": {"oid": "bbf28f09f2cb756a4ae02d511f77091a996fb812", "messageHeadline": "Fix a build warning: a constexpr function calls a non-constexpr funct\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "58af36b49a192b06cc9416a58ed79b429f020d7a", "messageHeadline": "Fuse ScaledSum and its backward BatchScale (#16517)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1595, "deletions": 4}}, {"node": {"oid": "c11ed065bacf8e7a31befa19ee9fd5c098ae02dd", "messageHeadline": "Fix SkipLayerNorm fusion in transformer optimizer (#17320)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 307, "deletions": 34}}, {"node": {"oid": "507a40e1e9b66494c1d14a05c83414db1c8362c8", "messageHeadline": "Add compiler cache in Linux GPU TensorRT CI. (#17348)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 48, "deletions": 36}}, {"node": {"oid": "47fe7fe90091e66fce7c6ef46d298ce610de359a", "messageHeadline": "Enable QDQ node unit support for Log op (#17354)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "64f06d0b4a9e5b3b97538c7cd809531a8f7b93af", "messageHeadline": "only Flush once for the same stream in copyInputAcrossDevice() (#17303)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "70e8c23944d8d0bf783fd209c229762d508b7fac", "messageHeadline": "[WebNN EP] Fix bug in interpreting ONNX's pads into WebNN's padding (\u2026", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 10, "deletions": 3}}, {"node": {"oid": "6c39641ea2248f616f1da306fe7b2d4798b321a8", "messageHeadline": "Fix a memleak in RunAsync python (#17326)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 24, "deletions": 6}}, {"node": {"oid": "081c0692a41fe92bafbfb2c1105dbac7b687bce0", "messageHeadline": "Update to nodejs version from 16 to 18.17.1 (#17351)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 12, "deletions": 8}}, {"node": {"oid": "2da08c477aedc62fd9b700bd005594e0dd130f49", "messageHeadline": "Add website publish placeholder (#17318)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 16, "deletions": 0}}, {"node": {"oid": "71da0824f3644e378cb2a70ce63f6e4e24044804", "messageHeadline": "Upgrade binskim and fix an error in nuget packaging pipeline (#17340)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "21ae86e4051751741ad9b92512595896853721b5", "messageHeadline": "[QNN EP] Fix test zero-point calculation and flaky MatMul test (#17338)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 90, "deletions": 28}}, {"node": {"oid": "922629aad81591be814e5c7d58475a392294b6e5", "messageHeadline": "Upgrade Centos7 to Alamlinux8 (#16907)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 97, "deletions": 89}}, {"node": {"oid": "c961f67b5ee5d433a4bf73554a196af021d6c12a", "messageHeadline": "Handle dtype attribute in float16 conversion script (#17321)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 50, "deletions": 29}}, {"node": {"oid": "8224891236ae612a3e6a59ea3420b944f54fae4f", "messageHeadline": "add logits option to generate artifacts (#17276)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 48, "deletions": 0}}, {"node": {"oid": "f3682eee3b89e73b447517445503b80664bca73d", "messageHeadline": "Fix log color, otherwise, the immediate line followed by the colored \u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "c438360c1e51450a071549cd9c208211cde49d02", "messageHeadline": "Noticed a simple simplification in beam_search_topk (#17275)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 4}}, {"node": {"oid": "d4a61ac71f35671358712890dc61e83019b29e30", "messageHeadline": "Pr trggiers generated by code (#17247)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 314, "deletions": 44}}, {"node": {"oid": "fd0917b27b9b9886695e515db785fd1274417d21", "messageHeadline": "Propagate noop_with_empty_axes in reduce operators. (#16845)", "author": {"name": "AtanasDimitrovQC", "email": "128688806+AtanasDimitrovQC@users.noreply.github.com", "user": null}, "additions": 246, "deletions": 11}}, {"node": {"oid": "7b920573760ff8a61bdbde01d1b965e895530bb1", "messageHeadline": "EvalStep called with wrong inputs onnxruntime_training_cxx_inline.h (\u2026", "author": {"name": "kushalpatil07", "email": "44136439+kushalpatil07@users.noreply.github.com", "user": {"login": "kushalpatil07"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e5ca3f3dcb1ae6cdc5d80b3776c5b70ec6354e4c", "messageHeadline": "[js/api] introducing IO binding for tensor (#16452)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 843, "deletions": 270}}, {"node": {"oid": "8827363fd2badf21fe84ef326ef033f27cbdda97", "messageHeadline": "Bugfixes: dangling pointers and python property typo (#17285)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "fffefb1c22a5c93d53511454bed844e9179beb0b", "messageHeadline": "[js/webgpu] Optimize matmul (#16969)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 311, "deletions": 94}}, {"node": {"oid": "4880f1da46e08008aa2f0c17df8cc79b4e40fdc6", "messageHeadline": "Fix attention fusion for UNet onnx model export when using LoRA weigh\u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 673, "deletions": 23}}, {"node": {"oid": "761c4333b5bd1ff36145e6aeb2349b0865a68c6b", "messageHeadline": "[QNN EP] GridSample op support (#17317)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 275, "deletions": 98}}, {"node": {"oid": "742b192a3414490fe4ab3f206e41942037acc774", "messageHeadline": "[QNN EP] Enable GlobalMaxPool op (#17304)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 341, "deletions": 267}}, {"node": {"oid": "6e60dba72645f146a0c1dd1b525ba620b368c51c", "messageHeadline": "Fix compilation with newer flatbuffers (#17164)", "author": {"name": "Artem Shilkin", "email": "89970996+reshilkin@users.noreply.github.com", "user": {"login": "reshilkin"}}, "additions": 18, "deletions": 60}}, {"node": {"oid": "0e9e9b2a67c4f96ab643216376883a7739fcaee7", "messageHeadline": "Fix one exception in post merge (#17327)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5d2c57363f491142cf885459bab39bf7c79dbe11", "messageHeadline": "Sign CUDA Kernel (#17293)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 173, "deletions": 95}}, {"node": {"oid": "38ea8c3931ce6e06fa2bccce41bff78d16d9af69", "messageHeadline": "Increase max error tolerance for ConvTransposeGrad test (#17315)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ee9d0461129005e4b9bb6ad6de412c9734aa410f", "messageHeadline": "Fix model serialization with external data in current directory (#17311)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 57, "deletions": 1}}, {"node": {"oid": "228db2431785afd0244e156210bfc6d0af24c1da", "messageHeadline": "Add training API functions to WASM API (#16521)", "author": {"name": "Caroline", "email": "wolfivyaura@gmail.com", "user": {"login": "carzh"}}, "additions": 299, "deletions": 11}}, {"node": {"oid": "cbd97515cd6566f1cd369d49240e5331c9028775", "messageHeadline": "[JS/WebGPU] Support GatherElements kernel (#17243)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 418, "deletions": 3}}, {"node": {"oid": "53169f59e5a34863de4036989ce3e84776e5f192", "messageHeadline": "[ROCm] Sort candidate solutions in rocBLAS/hipBLASLt for deterministi\u2026", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 40, "deletions": 20}}, {"node": {"oid": "bf8b1681f946a6556933f936476fa4863e883696", "messageHeadline": "Build nuget pkg for ROCm (#16791)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 408, "deletions": 6}}, {"node": {"oid": "bb1871332f5e37ebaa6a508fed460ab836fb23c5", "messageHeadline": "[js/webgpu] add kernel Not and Equal (#17306)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 47, "deletions": 40}}, {"node": {"oid": "4eedd3bb46ad596aa4167aa07c68c0c7edd7ba4b", "messageHeadline": "[TRT EP] Fix logic to reach cache encryption code. (#17111)", "author": {"name": "simonjub", "email": "78098752+simonjub@users.noreply.github.com", "user": {"login": "simonjub"}}, "additions": 39, "deletions": 22}}, {"node": {"oid": "ca0159b45d343a60df69418e6a49cb19bb1161b6", "messageHeadline": "Various test infra updates from testing Azure ops with MAUI test app \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 53, "deletions": 14}}, {"node": {"oid": "ddcd46174e9566163faff8cb38d7060f31d9cd1c", "messageHeadline": "[js/webgpu] fix jsepOnRunEnd (#17300)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 31, "deletions": 18}}, {"node": {"oid": "808215366d1aedb0567994c2f11acb2b9b6cbc24", "messageHeadline": "Fix Multi GPU TensorRT tests (#17269)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 50, "deletions": 90}}, {"node": {"oid": "c26287921456c0f99665c3d9c85eb0ab0b5347c7", "messageHeadline": "Added DML and CUDA provider support in onnxruntime-node (#16050)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 292, "deletions": 27}}, {"node": {"oid": "873ef8b8f0b09b49c0a7b7e2f03f3639d7418c22", "messageHeadline": "[js/webgpu] add label for some webgpu APIs (#17291)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "5e8d94cec8fc3554791243851e3cadf005ffa04c", "messageHeadline": "[js/webgpu] support Greater and Less operators (#17296)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 111, "deletions": 52}}, {"node": {"oid": "5a83a67f32da784b7dd4029bb23e7cc05363b748", "messageHeadline": "Support QDQ transformations with com.microsoft.Quantize/Dequantize op\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1656, "deletions": 917}}, {"node": {"oid": "79c4ed9a45c81d5fa71847789ccecb554ffdf76e", "messageHeadline": "[js/webgpu] support error pop and kernel name (#17260)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 112, "deletions": 40}}, {"node": {"oid": "da180b20fad36cdde0038d34a44b432927c96253", "messageHeadline": "[JS/Web] Fix ConvTranspose shader code compilation errors. (#17232)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 378, "deletions": 98}}, {"node": {"oid": "401129d4848c302eff995f08725ca9aa5daef15c", "messageHeadline": "Add support for more ops for padding elimination (#17217)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 65, "deletions": 19}}, {"node": {"oid": "735cc8e6c810f3738b0879470e14fd866069b374", "messageHeadline": "[ROCm] enable If op for ROCm EP. (#17279)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 13, "deletions": 10}}, {"node": {"oid": "9cd33e07b44d8fa0b1cb1664eaf687fc16705a91", "messageHeadline": "Readd Tests in Window GPU Reduced Ops workflow (#17294)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "4a0f8f6672d0d74b9c6420a3e7bb03c2889821b1", "messageHeadline": "Skip one flaky Test (#17290)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "3e934030f48bffc07c92e97a35b33cf2aad2d51f", "messageHeadline": "nodejs: Release Ort Env before main function returns (#17288)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 12}}, {"node": {"oid": "93ae17d1bb9eb81e781f626a7f07dde8529a761c", "messageHeadline": "[ROCm] Add hipBLASLt workspace support (#17096)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 27, "deletions": 28}}, {"node": {"oid": "7c98f45928de88726bacbd610ae0ab4ceffdc848", "messageHeadline": "Fix layernorm and softmax axis after upstream (#17255)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 405, "deletions": 7}}, {"node": {"oid": "86238fb5071ec668cbd0084992e3f408f4ab23fa", "messageHeadline": "[Docs] Auto generate JS API (#17271)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 50, "deletions": 101}}, {"node": {"oid": "756eda2cc499cb56640d4b7320ebdaee2fd3751a", "messageHeadline": "Windows CI build steps template (#17263)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 124, "deletions": 77}}, {"node": {"oid": "680fac64ed15c3ff3ad3f34b445c5861e59980fa", "messageHeadline": "[QNN EP] Support non-quantized Op on HTP (#17194)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 824, "deletions": 420}}, {"node": {"oid": "18d5cfdb858cf32ca14a714c9eeb4032683a685b", "messageHeadline": "Fix build - redefinition of default argument for \u2018long unsigned int E\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d90afc697be8fced404771f0eb64da8ce54d467d", "messageHeadline": "Introduce ZeROOffloadSubscriber for ORTModule (#17006)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1041, "deletions": 323}}, {"node": {"oid": "fca81cc5d5672dca039bece842f785b38bbd8f4f", "messageHeadline": "ConvTransposeGrad CUDA Kernel (#17201)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1475, "deletions": 268}}, {"node": {"oid": "33415b9da499702528f7dd3db8a1551f6108208e", "messageHeadline": "Removing 10.14 suffix  from osx nuget package (#17277)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "94f23882f710fbee5f2f9b3b7946906284fa1e1a", "messageHeadline": "Colorize terminal log output (#17196)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 51, "deletions": 0}}, {"node": {"oid": "34d18ee0762aa64ef32805488749f7181999a0cf", "messageHeadline": "Build gradient graph starting at the loss alone (#17240)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "fb51faea64b800d3df471472fb606d5fec0c765b", "messageHeadline": "[js/webgpu] fix 2 build breaks introduced in merge (#17273)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "87bef1f3f269ced5213faa8e24b3a4e2cb9c5e1f", "messageHeadline": "Move composable_kernel to deps.txt (#17245)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 16, "deletions": 8}}, {"node": {"oid": "33c87f62835e5605a82325f5d760be650586d726", "messageHeadline": "ORT_ENFORCE on the iterator must come before iterator is dereferenced\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "6c95d959f347c1b81c8057c6ce1666468407986d", "messageHeadline": "Make batchnorm training mode available in inference only package (#17\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 12, "deletions": 7}}, {"node": {"oid": "fdc3bcae207a7582cbdaace8e5ad37eb85c6ddf2", "messageHeadline": "Disable local symbol table for function shape inferencing. (#17267)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "8b18d48c7c1ac2dde474dc69c5c1f2e4510cf316", "messageHeadline": "[js/webgpu] make IndicesHelper implementation implicit (#17193)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 98, "deletions": 114}}, {"node": {"oid": "aed7c6ffc757245a9b0955a3f11a9b95a3c69de4", "messageHeadline": "Exclude fp16 support flag definition from minimal build (#17259)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "b3cb775cf9076321deaf7a6eb284fbc2da68adb8", "messageHeadline": "Two fixes involving minimal builds (#17000)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 146, "deletions": 26}}, {"node": {"oid": "d21a2f064b3307bc7ae29ba26bbb1d0cf1c45bf8", "messageHeadline": "[VITISAI] fix compile error for onnxruntime (#17252)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "56102ecbdd5d46a565dbde1e4f69c9083122161b", "messageHeadline": "On-Device Training - Enable loading from buffer (#16417)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 524, "deletions": 140}}, {"node": {"oid": "ae62d752d60b656a7023518f77b83040fc14ddb4", "messageHeadline": "Prevent GSL_SUPPRESS arguments from being modified by clang-format (#\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 53, "deletions": 59}}, {"node": {"oid": "4b3477f1710da3a2f74877db65dffc6758b7ed74", "messageHeadline": "Add Whisper scripts (#17043)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 1138, "deletions": 14}}, {"node": {"oid": "5842144d987bb1ec336441b2decf20914652c318", "messageHeadline": "[js/web] JSEP Gemm for opset 13 (#16936)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 14, "deletions": 4}}, {"node": {"oid": "edac3ef150d278f79440d7bd5668b1e7da81c21c", "messageHeadline": "Add LLaMA scripts (#17020)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 2035, "deletions": 1}}, {"node": {"oid": "d3d3dde844029363aed6edcefe59ea3efc77dcfe", "messageHeadline": "fix webgpu split (#17258)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "d76dbc4fc31abc18c09c42e7c254e1ac8a88f723", "messageHeadline": "Issue#16990: Cast -> AllToAll -> Cast fails with random output (#17075)", "author": {"name": "shaahji", "email": "96227573+shaahji@users.noreply.github.com", "user": {"login": "shaahji"}}, "additions": 163, "deletions": 55}}, {"node": {"oid": "bd8a488f4b6d1ad3b3ce9ad0bc1e772d136a805c", "messageHeadline": "Enable verbose logging in unit test program with environment variable\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 29, "deletions": 3}}, {"node": {"oid": "61a79436e22892bdd91a905389f12e0aee68132e", "messageHeadline": "Common pre-build steps of Windows CI (#16970)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 131, "deletions": 141}}, {"node": {"oid": "d5c565156db023af400ff4e18780bdb2862eb343", "messageHeadline": "[ROCm] add SimplifiedSkipLayerNorm implementation (#17213)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 249, "deletions": 84}}, {"node": {"oid": "4e6cec4d09ca399c66541ee61109c3099af1a463", "messageHeadline": "Update ck and enable test (#16383)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 54, "deletions": 44}}, {"node": {"oid": "aae9a52e8b4d3c650cace82c95b2c6ff1b6c02d6", "messageHeadline": "Avoid pushing cpu package to https://download.onnxruntime.ai/ (#17238)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 0, "deletions": 28}}, {"node": {"oid": "ced0cfbfea9d0ad4ec85349e4397e837da2e33b7", "messageHeadline": "[C#]Fix API Comment (#17236)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "618f4839d16b8dff90ba4d2c593a8e10de325d0d", "messageHeadline": "[QNN EP] Re-enable some node tests for QNN (#17237)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 0, "deletions": 14}}, {"node": {"oid": "e2b6827a590d8199017929d0c15e3e065058b23e", "messageHeadline": "Add a CUDA 12.x pipeline and improve install_third_party_deps.ps1 (#1\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 227, "deletions": 53}}, {"node": {"oid": "08ca624d2b65142218a5128bd20787cecb4e1591", "messageHeadline": "Fix: update hyperlinks to the Jupyter notebooks (#16145)", "author": {"name": "Emmanuel Ferdman", "email": "emmanuelferdman@gmail.com", "user": {"login": "emmanuel-ferdman"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "cbaa00839177650073da298d7693e7e42f6940e1", "messageHeadline": "Bump DirectML version from 1.12.0 to 1.12.1 (#17225)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "4bea5ec513f3fea988ca11c4454ea48d575bbc12", "messageHeadline": "Add Whisper export with beam search test cases (#17228)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 112, "deletions": 2}}, {"node": {"oid": "9445539e2cdd6cd11e67020c67086fb2c822214f", "messageHeadline": "Update dependency for deps.txt (#17220)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "6fc3fd9ece6e21ad42d8714172e3eee1742cde10", "messageHeadline": "[js/webgpu] support Cast operator (#16489)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 393, "deletions": 10}}, {"node": {"oid": "bf1c62c18172fcb5851a47b2d05b2ee2360bdfff", "messageHeadline": "check in build script for webgpu (#17126)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 91, "deletions": 0}}, {"node": {"oid": "d6cd41cfc19dd9cc4897710346aea79133b511ab", "messageHeadline": "[CoreML EP] Add Shape, Gather, and Slice ops (#17153)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 984, "deletions": 432}}, {"node": {"oid": "2b4cc24d5cdece3b6ce53130cb2c97c664cdf335", "messageHeadline": "[CoreML EP] Limit input shapes to at most rank 5 (#17086)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 154, "deletions": 95}}, {"node": {"oid": "34269545255b6663a2fb5b2b86313085f2c45fd1", "messageHeadline": "disable browser stack tests (#17224)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 10, "deletions": 7}}, {"node": {"oid": "3cec88bd12708fc56d4a56e95a5398b15a2956b4", "messageHeadline": "FIX: memory leak checker is incompatible with std::stacktrace (#17209)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 4}}, {"node": {"oid": "6db72165eb3269e768f0fea28a7b210f588c7abe", "messageHeadline": "Fix python packaging test pipeline (#17204)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 105, "deletions": 43}}, {"node": {"oid": "dd3b2cefd6abdd6a0fe6f6701d26d1299f87f1fa", "messageHeadline": "[js/webgpu] Support int32 type for binary (#16901)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 113, "deletions": 41}}, {"node": {"oid": "c0b6c6c94bcd7cb1ed6bb0fc8edfa8b606229411", "messageHeadline": "Add SGDOptimizer in the on-device training offline tooling (onnxblock\u2026", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 266, "deletions": 123}}, {"node": {"oid": "ee09a5ff35b4424276d4b382c386a350092c3486", "messageHeadline": "Add DISABLE_CUSPARSE_DEPRECATED flag to CUDA build (#17207)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "a476dbf430ac8315550474a78d47bf182f202d7c", "messageHeadline": "[JS/WebGPU] Support Tile operator (#17123)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 305, "deletions": 2}}, {"node": {"oid": "7d1a5635a01ed9294988bd46211f84a1cb80df44", "messageHeadline": "[JS/Web] Added SkipLayerNormalization operator. (#17102)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 388, "deletions": 4}}, {"node": {"oid": "9266cf1772ca45c2a5b5b5596014576d98f9736f", "messageHeadline": "Skip setting the name when AzureEP enabled. (#17208)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "68a670c7f8ae05f83b88209f6a7e58e35518c429", "messageHeadline": "Move some tests from CUDA only to CPU (#17189)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 35, "deletions": 23}}, {"node": {"oid": "d65aa5400c9fa9669c428d737e386c22710f2874", "messageHeadline": "clean up transformers scripts (#17179)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 182, "deletions": 251}}, {"node": {"oid": "78b35652a3ada2a08ea058c97a34d6e9ad786a39", "messageHeadline": "fix issue with obtaining the decoder layer number when converting the\u2026", "author": {"name": "Jack", "email": "lityangweiguang@163.com", "user": {"login": "LitLeo"}}, "additions": 18, "deletions": 12}}, {"node": {"oid": "6ee4be724bab79377b4da3133d0a50d0fd7a284c", "messageHeadline": "Update LICENSE name in NuGet packaging pipelines (#17183)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "5c54b64a63e4c3a8da5635756e3bd8582533572d", "messageHeadline": "Create NodeArgs for all Constant nodes and initializers for functions\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 107, "deletions": 85}}, {"node": {"oid": "0cccbcc47b46a842dc7c72c3ac463fe9187be873", "messageHeadline": "Move DML build job's Prefast task to a CPU machine pool (#17192)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 84, "deletions": 37}}, {"node": {"oid": "e0022d061f84e86c21e38273a260aa3ac3f70185", "messageHeadline": "Set web-ci-pipeline.yml only triggered when related fields are update\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 75, "deletions": 28}}, {"node": {"oid": "df124c9313a510575a9bf4590704cf10feef0fa6", "messageHeadline": "[VITISAI] 1. Fix reading .dat and .onnx on Linux 2. Fix issue of comp\u2026", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "2fb148dd88a4f017686a5c8e863829ab87b4ea58", "messageHeadline": "Temporarily enforce \"Debug build\" TRT EP with trt oss parser on Windo\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 9, "deletions": 2}}, {"node": {"oid": "59a28011368fcd23351dd270b1c404220553a8f9", "messageHeadline": "Fix NuGet pkging pipeline (#17195)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "049adb9f31c6823977fb4e3eaecf92e5ebb7ab75", "messageHeadline": "[ROCm] Remove redundant ep field in softmax (#17048)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 4, "deletions": 22}}, {"node": {"oid": "5249b7ab7cee8fc6f58e7a448552aa68e11f7000", "messageHeadline": "Re-implement stacktrace (#17173)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 99}}, {"node": {"oid": "f45eef399ea68f88ac680390d2615da7e4f55f5a", "messageHeadline": "Fix visualization issues with Attribute/Tensor protos (#17188)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "3dd2c1b4d7e40a35495cd009683ea606674c90c3", "messageHeadline": "EP context for custom op (#16454)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 669, "deletions": 247}}, {"node": {"oid": "7b9d1f18c7d39314b369b7c56bae22fafe897405", "messageHeadline": "NVTX windows include and link fixes (#16831)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 13, "deletions": 3}}, {"node": {"oid": "cbee84ddfb68b50246ce76ce59338526b2f65bde", "messageHeadline": "[js/web] allow optional input/output in operator test (#17184)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 190, "deletions": 82}}, {"node": {"oid": "96b1ff610bf5e31c266fcc6d295307a26f307718", "messageHeadline": "Add CI and PR validation triggers to QNN Windows x64 Pipeline yaml (#\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 29, "deletions": 0}}, {"node": {"oid": "66df11769c127565dcd17a715b2e08b3bffd9eee", "messageHeadline": "[JS/WebGPU] Expand operator fixes (#17137)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 93, "deletions": 3}}, {"node": {"oid": "99349e58d7127ae8ce268c31b299aeb75a941c42", "messageHeadline": "dump tensor statistics (#15761)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 163, "deletions": 1}}, {"node": {"oid": "89b682e3f3dc90c5f85dcab92cc140744fc4cc4f", "messageHeadline": "[JS/Web] The bias input is optional, not required, for LayerNormaliza\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 51, "deletions": 9}}, {"node": {"oid": "2ae930333b0624d57dfaaf8cc216128453a58c28", "messageHeadline": "Add checks for session options and fix gsubgraph fallback exceptions \u2026", "author": {"name": "Preetha Veeramalai", "email": "preetha.veeramalai@intel.com", "user": {"login": "preetha-intel"}}, "additions": 31, "deletions": 11}}, {"node": {"oid": "133af1385cc08da45d46b5da673942a4ba6bb27f", "messageHeadline": "[js/webgpu] update shader cache key to include input tensor datatype \u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "8998b6811dd654c84c773d447807265b82acfa9d", "messageHeadline": "Fix NPM Packaging Pipeline (#17182)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ebcd9b5cae2ce221cfdb080eb9b9a69b48557f78", "messageHeadline": "Fix deprecated optimum interface (#17112)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "6b29837ed28bb550a8f4b09f888225eea446c0f6", "messageHeadline": "Move attention test data to file (#17158)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 5739, "deletions": 6472}}, {"node": {"oid": "33ecde9af1cc38634f6431d2f1124502a46e20c9", "messageHeadline": "[js/webgpu] Fix reshape int32 test case (#17113)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 96, "deletions": 1}}, {"node": {"oid": "8289e8b6efab41db766fd22b61f1a1f82c742834", "messageHeadline": "[js/webgpu] fix a few shader errors (#17171)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "35363dd9a5ebdd0d58ecdca6f4b18229f8ae5d99", "messageHeadline": "[js/web] a few optimizations for test runner (#17174)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 8, "deletions": 6}}, {"node": {"oid": "2575b9aaa156c7ec4e4e6ca15b405c6fd6ae1b89", "messageHeadline": "Improve comments in winml/ (#17163)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 91, "deletions": 118}}, {"node": {"oid": "178e5991ac8c03c5764addd3bf04469a4a691d84", "messageHeadline": "Bump protobufjs from 6.11.3 to 6.11.4 in /js/node (#17177)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "ccf14e891e91900c9336b3444c4618f655d4fffd", "messageHeadline": "[js/web] JSEP node assignment optimization (#17128)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 51, "deletions": 26}}, {"node": {"oid": "3cdf42548fcccaee05ce5a5e0f9ee563105c8665", "messageHeadline": "Issue #17098: Shape inferencing fails during quantization for large m\u2026", "author": {"name": "shaahji", "email": "96227573+shaahji@users.noreply.github.com", "user": {"login": "shaahji"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "789bac1dc8d0796f57c6f2730d525e8fb93f32e8", "messageHeadline": "[WebNN EP] Support BatchNormalization op (#17071)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 72, "deletions": 48}}, {"node": {"oid": "c0f8197157dfa609d97bfc5e2ea0d66527f29d99", "messageHeadline": "Add README to Nuget and fix license file name (#17170)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "39dfcd5d84d428f7e3c7fd99a9a74540642b1f5d", "messageHeadline": "Allow RunAsync with global TP (#17157)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 71, "deletions": 11}}, {"node": {"oid": "c647e3e8abf6b7ba1a55ace2b84ff595c36eae0a", "messageHeadline": "Run nightly pipeline tests from the commit id. (#17162)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "f086bd7bff850d7accfc679301420cc16bc0c083", "messageHeadline": "Bump github/issue-labeler from 2.5 to 3.2 (#16639)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8e203efc69c30cb43fd61f3d5efefe0273d4ad19", "messageHeadline": "Cleanup cmake file (#17154)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 45}}, {"node": {"oid": "2a22325005e2001227464119991101911ecd296a", "messageHeadline": "Explicitly set JDK version when building ORT java package (#17147)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "90754fc077a13a377ca00fc762318a6280e33f61", "messageHeadline": "Fix invalid escape sequence (#17145)", "author": {"name": "Severin Simmler", "email": "s.simmler@snapaddy.com", "user": {"login": "severinsimmler"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "3aba736ee2bf3228a95c823f20dbfce0c3b1ff61", "messageHeadline": "Refactoring of Stable Diffusion scripts (#17138)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 594, "deletions": 716}}, {"node": {"oid": "5e971bc51a6901a8fb968067e25bccf0e7f6a728", "messageHeadline": "Rework WIL dependency retrieval/usage (#17130)", "author": {"name": "Matthieu Darbois", "email": "mayeut@users.noreply.github.com", "user": {"login": "mayeut"}}, "additions": 21, "deletions": 27}}, {"node": {"oid": "412b0d0831a0c8da16ac0304313f6942717aabe6", "messageHeadline": "Update BERT and GPT-2 optimization notebooks for CPU EP (#17057)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 356, "deletions": 318}}, {"node": {"oid": "abf9765d733b88800cac58a60b64f32ab3744e0a", "messageHeadline": "PythonOp Enhancement: Bool and Tuple[Bool] Constants, Materialize Gra\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 840, "deletions": 284}}, {"node": {"oid": "b734db1924a2f9e0dfe2b9ae44679cccbf171c7f", "messageHeadline": "[QNN EP] Fix CI build on Windows x64 pipelines (#17152)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 20, "deletions": 19}}, {"node": {"oid": "416dc2e84d04835827a3b52eb5872716ec04e708", "messageHeadline": "Fix clang-format comment indents on Windows for winml/ (#17144)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 304, "deletions": 304}}, {"node": {"oid": "24e0bd37b41671b9d8659c761865e7fd63a37b6b", "messageHeadline": "[JS/WebGPU] Support Log operator (#17045)", "author": {"name": "xhcao", "email": "xinghua.cao@intel.com", "user": {"login": "xhcao"}}, "additions": 15, "deletions": 1}}, {"node": {"oid": "289600b47dfcf837365d672a3a69ed823f989163", "messageHeadline": "ONNX Runtime training cpu package name for ADO (#17109)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "be2200c00b0101e51bd4ffc4644a540d5c573773", "messageHeadline": "[ROCm] fix python package pipeline (#17136)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 0, "deletions": 16}}, {"node": {"oid": "45f52987a2fc9445effe6cf9e89a235c4501ca8b", "messageHeadline": "Web CI Pipeline Isolation (#17005)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 540, "deletions": 8}}, {"node": {"oid": "d052c8a45c639b9c51be37b9b06f97e0c4d0d02c", "messageHeadline": "Remove the extensions submodule (#17097)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 1, "deletions": 12}}, {"node": {"oid": "68ea9631af0e9610407950bb32a597cba4f841e5", "messageHeadline": "Fix typo onnxruntimecpubuilpython (#17120)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f71b6944bffa8deccc63206ce0b88c2ffc34a063", "messageHeadline": "Fix nuget pipeline (#17110)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "e55e1b7da977e73dec5d31740aab65f28c450a72", "messageHeadline": "Mark end of version 16 C API (#17107)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "cd7b3f54da8b188d0cc52ddc026df086a827c39f", "messageHeadline": "Allow defining customized PythonOp shape inferer (#17093)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 195, "deletions": 24}}, {"node": {"oid": "9204cd73920b725219719eb4162b9dc04b02f170", "messageHeadline": "[js/webgpu] Add C++ registration for operator Tanh in JSEP (#17124)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "e7adbb38f69fb1dd9e73a8fd0cb0e532c71b0d41", "messageHeadline": "[js/webgpu] disable test case 'test_batchnorm_epsilon_training_mode' \u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f2e1b91634a42091db49cebf196f6d96788ba0d5", "messageHeadline": "add int4 quantization code in python (#17077)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 244, "deletions": 54}}, {"node": {"oid": "5704e71b8909f2fdb04dd3da0dc75a96ca4a261d", "messageHeadline": "update onnx.patch to apply wasm build break fix (#17104)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 15, "deletions": 2}}, {"node": {"oid": "6697635b9182abaf562d1f7aaafe011dfdc083dc", "messageHeadline": "To support size opset 19 (#15689)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 26, "deletions": 6}}, {"node": {"oid": "14a8315f10e13d384e2a810c43c582d3952b5b94", "messageHeadline": "[js/web] [webgpu] new incides helper (#16957)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 957, "deletions": 672}}, {"node": {"oid": "4728f20f9abfedab1102741e4255ca327df470d0", "messageHeadline": "Fix CI build (#17118)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "e7e974b23f7d2c9f4ae5ac4b5e5c8629c1ef1746", "messageHeadline": "Use double quotes so variable gets expanded. (#17105)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ebaeda6c23cb84b101bbafaed1a8a39f5c7a9bc3", "messageHeadline": "try to find patch.exe in git default installation folder (#17106)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "54153c73f0d26fbefaa84d864c9adc481c9d88cd", "messageHeadline": "Batchnorm training mode support in a minimal build (#17103)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 7, "deletions": 12}}, {"node": {"oid": "344c41fdb96f00b0319ccd0b9086f05056537e27", "messageHeadline": "[QNN EP] Update QNN to v2.13 (#17079)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 63, "deletions": 48}}, {"node": {"oid": "3e7f70bf8858c9dffde45be4f4670e9ffd71a7eb", "messageHeadline": "LeakyRelu Gradient (#17039)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 183, "deletions": 2}}, {"node": {"oid": "0180c0429f6f4ac529f9fa5feca840382ce1c393", "messageHeadline": "Fix DML regression from allocator refactor and enable unrounded weigh\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 19, "deletions": 89}}, {"node": {"oid": "9cd4e5af688cda3d13b8a16379f3c9d1792d1d7e", "messageHeadline": "[wasm] upgrade emsdk to 3.1.44 (#17069)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 79, "deletions": 137}}, {"node": {"oid": "66b45e00858473c74fa16c7116f6dcdcb7b7642f", "messageHeadline": "Bump actions/upload-pages-artifact from 1 to 2 (#16727)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "83240d1346f750ee196a2579d423d4e777d481f0", "messageHeadline": "Bump clang-format to 16.0.6 in CI (#17099)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6986981482086832d897a9b3241987c04c0a64cf", "messageHeadline": "Bump ONNX version (#16325)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "6dffd1a890b2d439da7d5b79d60be16b69db4fb4", "messageHeadline": "Update model_tests.cc: avoid auto adding new tests from new opsets (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 154, "deletions": 121}}, {"node": {"oid": "12837ba5c7eecb38fffb85e23a6781bd8b9ebaac", "messageHeadline": "[ROCm] Update CI based on ubuntu 22.04 (#17076)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 175, "deletions": 63}}, {"node": {"oid": "87285323e683e7adc48cf3eab17a00ca0b4fe8e2", "messageHeadline": "[VITISAI] nested subgraph is unsupported for now (#17067)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "1b081d51dc5c17548bcc04be1f976640f7299333", "messageHeadline": "[VITISAI] node arg can be used more than once (#17068)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 8, "deletions": 11}}, {"node": {"oid": "e8a9d4f04d8dc49fba87d21d181d8e8c12dea5b7", "messageHeadline": "[JS/Web] Fix Resize kMSInternalNHWCDomain (#17023)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 49, "deletions": 35}}, {"node": {"oid": "ef6f4a4aa1dbf9278379e630055a3075e5cda60e", "messageHeadline": "support broadcast shape for elementwise node in padding elimination (\u2026", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 169, "deletions": 86}}, {"node": {"oid": "b4e0fc87ea746cf0bcb024e8f60c4d1e610bb2ae", "messageHeadline": "[ROCm] Make KE reports with better format (#17049)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 38, "deletions": 35}}, {"node": {"oid": "0471f6fbb3aa1602cfc3d90e2daa8c190970869b", "messageHeadline": "Check type for building gradient graph (#17046)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 71, "deletions": 20}}, {"node": {"oid": "31cbd63af7a8a8582865882b1870719724b6e9b4", "messageHeadline": "GRU Training and GRU Gradient Kernels (#16929)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2578, "deletions": 135}}, {"node": {"oid": "249c2221b69e3ac9d7bcd9e9d0455b59a4de1c8d", "messageHeadline": "[VITISAI] remove unused code (#17066)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 0, "deletions": 17}}, {"node": {"oid": "dbbfc249f73655d3745b3d708eb9d4242ad83366", "messageHeadline": "[ROCm] update header and binary search paths used by cmake (#17083)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 22, "deletions": 4}}, {"node": {"oid": "7c7c9914177834f889309db7b9b56278b4aedf43", "messageHeadline": "[ROCm] Workaround type conversion issue (#17074)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 22, "deletions": 22}}, {"node": {"oid": "7201dbebe53cde956a9294a47b74d61196065d0b", "messageHeadline": "[DML EP] Split fused kernels when the persistent resource is too big \u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 286, "deletions": 189}}, {"node": {"oid": "e951f837e43369a71663fe67beee8165397a93bd", "messageHeadline": "[VITISAI] fix out of bound error on graph with loop (#17065)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "f17efb5c7b9b2d801463b6d92d00b2231706ab6d", "messageHeadline": "Copy to buffer for both trainable as well as non trainable parameters\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 30, "deletions": 21}}, {"node": {"oid": "555f34692339d6504c195a81e6bec880e0f50061", "messageHeadline": "[QNN EP] Enable DepthToSpace & SpaceToDepth Ops (#17038)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 111, "deletions": 1}}, {"node": {"oid": "a3e02e8e2a80ce0b471cb454c82809e10ed67594", "messageHeadline": "Fix Resize op input check (#16594)", "author": {"name": "Zimon Tai", "email": "daizhuoxian@gmail.com", "user": {"login": "zimond"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "7d340256f1e6e94865fd5bb6614c0017e4538b40", "messageHeadline": "Add \"windows_sdk_version\" build arg and fix SCA build pipeline (#17062)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "d793e239b049fe60f6af70e3edc2def75865345c", "messageHeadline": "[QNN EP] Increase tolerance for ReduceProd test on x64 Windows (#17078)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 18, "deletions": 4}}, {"node": {"oid": "4bc2287a85205d682121fd2db0bdaaa2494615f4", "messageHeadline": "Fix GroupNorm tests failing when no providers are supported (#17054)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 25, "deletions": 21}}, {"node": {"oid": "a7542f48d65f427338a443d4533d353d54202f4c", "messageHeadline": "Make AzureEP default for python and c# packaging (#17025)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 68, "deletions": 130}}, {"node": {"oid": "2c5d4dce773ef9cdfb7e344046aaba9940d2a6ae", "messageHeadline": "Openvino ep ort 5.1 (#17042)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 333, "deletions": 211}}, {"node": {"oid": "03c3e91b0d6d5d76011ed0306d6b272a2f404ad7", "messageHeadline": "[java] Relaxing CoreML test (#16777)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "07dfe34714d18820a5ffda19cb118f7fecacbf1a", "messageHeadline": "Fix FunctionProto visualization (#17063)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 21, "deletions": 18}}, {"node": {"oid": "7361c283c7b32583cffd2fa1bb2e0e6038f2f354", "messageHeadline": "Add API for updating CUDA EP provider option user compute stream (#17\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 119, "deletions": 4}}, {"node": {"oid": "a4902ee65b2899a247ea85e07c69b253ef6fa11f", "messageHeadline": "[CUDA][ROCm] Allow allocating ScratchBuffer from TuningContext (#17028)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 281, "deletions": 203}}, {"node": {"oid": "6e6f582e0875f15837b338c402b065fd95e0dd46", "messageHeadline": "Use full qualified name for PythonOp export (#17021)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 157, "deletions": 139}}, {"node": {"oid": "c424e42594d92daba54f264c1c7409e53529d933", "messageHeadline": "[C++] Correctly handle scalar inputs in reduction ops, enforce Transp\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 53, "deletions": 9}}, {"node": {"oid": "fb11c673682cefe70992ecf6dc5c22289ebb24dd", "messageHeadline": "Fix SkipLayerNorm for 2D input (#17014)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 31, "deletions": 49}}, {"node": {"oid": "73037978f85d0a5d4b1f601101dbf65e04a43456", "messageHeadline": "Add PerThreadContext for TRT EP (#16599)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 653, "deletions": 273}}, {"node": {"oid": "56bced0581ea976f6fdd836fd6f7d5af4fd81566", "messageHeadline": "[js/web] enable webgpu in browser unit test (#16310)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 69, "deletions": 41}}, {"node": {"oid": "c3f04251c74c296da982caf683a16e06af2722d4", "messageHeadline": "[js/web] JSEP LayerNormalization and InstanceNormalizations kernels (\u2026", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 490, "deletions": 38}}, {"node": {"oid": "5b9bf8b6631b83aa48aecdd70b3a3faf4e060d73", "messageHeadline": "[TensorRT EP] Fix bug for using correct device id for EP allocator  (\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "50719d2f8e45d891cdb6e7fa3a5f711fe7b842d4", "messageHeadline": "[iOS] Add script to get simulator device info. (#17012)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 181, "deletions": 18}}, {"node": {"oid": "45ea907f530a16195436bfdc9a70f7b20ac99446", "messageHeadline": "Fix orttraining_test_dort.py (#17034)", "author": {"name": "Ti-Tai Wang", "email": "titaiwang@microsoft.com", "user": {"login": "titaiwangms"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "d0316ee7688bd8fd795751aaad9e926c409a1532", "messageHeadline": "Updating QDQ to support Float8E4M3FN (#16550)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 2479, "deletions": 232}}, {"node": {"oid": "063e9054b8056037c6c2af8de7acd0b66dadbac9", "messageHeadline": "RunAsync in C# (#16890)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 235, "deletions": 4}}, {"node": {"oid": "249917a093b85ec2234ae8b1ba904ab181a9a7d0", "messageHeadline": "Add mac and windows python packages for onnxruntime-training (#16993)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 71, "deletions": 36}}, {"node": {"oid": "e48dc3b281c60c905a021f70b609c389b129bbff", "messageHeadline": "Parallelize Transpose (#16854)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 22, "deletions": 21}}, {"node": {"oid": "3c10f027de52b57b2a28b8d4b5b1ee25173c7b81", "messageHeadline": "4b quantization for weights of LLMs (#16833)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 4763, "deletions": 17}}, {"node": {"oid": "8a335b8347f4a93d55e2fd5a090b3bf9dabb1376", "messageHeadline": "Update torch.onnx.OnnxRegistry usage in DORT tests (#17009)", "author": {"name": "Ti-Tai Wang", "email": "titaiwang@microsoft.com", "user": {"login": "titaiwangms"}}, "additions": 23, "deletions": 19}}, {"node": {"oid": "4e6ea730d633756e9e04df8968304d11a575dde4", "messageHeadline": "Broadcasting for SLN for CPU and CUDA (#16510)", "author": {"name": "Khalia Spear", "email": "134653808+khspear@users.noreply.github.com", "user": {"login": "khspear"}}, "additions": 276, "deletions": 117}}, {"node": {"oid": "3649376f09d238394cf0c22de14db3f4e8c11310", "messageHeadline": "Fix few small bugs (#17019)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 47, "deletions": 28}}, {"node": {"oid": "a451318820f76b878c6c766ada4dccbe0092670d", "messageHeadline": "Refactor TRT EP error message with details (#17007)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 16, "deletions": 2}}, {"node": {"oid": "d5e4bdbe7de2377130ae4953be1f30322ffcbf57", "messageHeadline": "Fix protobuf TaggedStringPtr display (#17008)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "78a5f049f416eb55d7984900bb6707a740e7221d", "messageHeadline": "[DML] Model corrupter during layernorm fusion and DmlNonZeroOperator \u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "d6ce43db5e4f5e2d6a09beeccc90c5fd22b4f009", "messageHeadline": "[EP Perf] MemTest: Add Valgrind and fix addressSanitizer (#16930)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 160, "deletions": 252}}, {"node": {"oid": "5af8774a0b1c6fb7f4afc0dedb6ed563db411672", "messageHeadline": "[build] do init and precheck first (#16961)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 59, "deletions": 47}}, {"node": {"oid": "fc8003349e2af25277ebc0213701d32db0b35726", "messageHeadline": "Add API for updating TRT EP provider option user compute stream (#16965)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 122, "deletions": 3}}, {"node": {"oid": "9ea0a3129b09462753a95c22d50655601367215c", "messageHeadline": "[js/webgpu] Make sure only storage buffers are reused (#16893)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 32, "deletions": 14}}, {"node": {"oid": "7ad43d9564a935f32c999d8e9ed8fcb30e437e7a", "messageHeadline": "[JS/Web] Fixed ArgMin and ArgMax and refactored (#17002)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 131, "deletions": 190}}, {"node": {"oid": "191f98a00ee4ec812b00c55b0cf371c28495a4d0", "messageHeadline": "[QNN EP] Improve QDQ model accuracy tests (#16916)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 2234, "deletions": 1589}}, {"node": {"oid": "e5bb7aba502f5a8783de945258d226c092c14386", "messageHeadline": "Add Gradient for Reciprocal (#16945)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 48, "deletions": 0}}, {"node": {"oid": "555414f1aa77613b834585a51322c16e7cc5ffe8", "messageHeadline": "Set PR trigger rules (#16987)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 103, "deletions": 0}}, {"node": {"oid": "a6887f171f09aae27250b6b0b27118a0f7628276", "messageHeadline": "Refactor schema extraction and output unflattening (#16894)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 718, "deletions": 266}}, {"node": {"oid": "f98d3f8a2340fa82e8530ada18c32175176c9aa9", "messageHeadline": "[CoreML EP] Enable inputs with dynamic shape (#16915)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1065, "deletions": 418}}, {"node": {"oid": "1629a6fa75ee28987d4d7b8db2d3fe465768ac1d", "messageHeadline": "[ROCm] add gfx1100 and gfx1101 to CMAKE_HIP_ARCHITECTURES (#16972)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cc4b64f646a9ae2ae79583f27a0f89769e70277a", "messageHeadline": "[JS/Web] Modify Reduce, Expand and Slice to pass op and node tests. (\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 64, "deletions": 54}}, {"node": {"oid": "a25d0d296b1a5d2677c34b3ca8c06444b86879fc", "messageHeadline": "Add --mask_type option to generate different format of attention mask\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 90, "deletions": 16}}, {"node": {"oid": "bda012a4b23ad7f94df02dd36714321956c21d4f", "messageHeadline": "Scripts to convert model with MulitHeadAttention to packing mode (#16\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 275, "deletions": 73}}, {"node": {"oid": "06096fcb31c9446e7951b87eac770998a90c54ed", "messageHeadline": "Hardcode xcodebuild destination iOS simulator OS to 16.4. (#16982)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "641c3a4a37e56944bf9d2a22915f3d562c3daf8f", "messageHeadline": "[js/web] update op test schema (#16921)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 510, "deletions": 67}}, {"node": {"oid": "ea55700e1cb3e4f8485e9c9cfd69ebe470701397", "messageHeadline": "[js/web] JSEP Gather OP (#16855)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 199, "deletions": 3}}, {"node": {"oid": "acb9e5616433cdb2af0469f54a1e59c3c355893c", "messageHeadline": "[js/web] JSEP Expand fix for inputs with rank < 2 (#16829)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "757c42cea7ce0368ace32f30991166cea4972411", "messageHeadline": "[rn] Update expo/config-plugins to 7.2.4 due to security warning with\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 39, "deletions": 41}}, {"node": {"oid": "c11cffb56543278b4d894df8dcce8b60bb927fab", "messageHeadline": "[js/web] Fix typo in JSEP ConvTranspose (#16884)", "author": {"name": "Arthur Islamov", "email": "arthur@islamov.ai", "user": {"login": "dakenf"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e6c9ed06066916d622df6a1b1bd415a4dbf829c3", "messageHeadline": "More element types in AllGather and AllToAll (#16941)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 209, "deletions": 52}}, {"node": {"oid": "b8bbc898c6436076f2852a461cd273ecf2bca805", "messageHeadline": "fix errors for node with empty name for vitis ai (#16949)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 4, "deletions": 10}}, {"node": {"oid": "246cb3a197007ecc632f26550138f501dd890979", "messageHeadline": "Simplify shrink, replace Eigne in Sign implemenation (#16975)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 14, "deletions": 39}}, {"node": {"oid": "0df2e14038394dc3cef3de6384ff2c7514d1e187", "messageHeadline": "js/webgpu: argmax,argmin,softmax support (#16882)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 544, "deletions": 47}}, {"node": {"oid": "506ddb3d5d6ed6b91635928ffde0fb28588162e2", "messageHeadline": "[js/WebGPU] Support int32 Transpose in WebGPU (#16952)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 64, "deletions": 6}}, {"node": {"oid": "6361b221033f04df2b38fded82030b89fd73ffa0", "messageHeadline": "vitis ai support generic data type (#16902)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 8, "deletions": 27}}, {"node": {"oid": "d39964886987313ec3659979c4f6375fd1a1a74c", "messageHeadline": "[JS/Web] Added Resize kMSInternalNHWCDomain domain registration. (#16\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 62, "deletions": 65}}, {"node": {"oid": "07e6648e1230721b9d2c8ed298719a12de9e1d33", "messageHeadline": "Enable Intel oneAPI DPC++/C++ compiler build (#16587)", "author": {"name": "Michael Klimenko", "email": "mklimenko29@gmail.com", "user": {"login": "MKlimenko"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "76aff63f375be7e295968c3ea0eca932973a4ae0", "messageHeadline": "Update bert_perf_test to test inputs with different padding ratio (#1\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 122, "deletions": 21}}, {"node": {"oid": "c392fdeb1b442aa213b664edc55e079efb336540", "messageHeadline": "RunAsync Python API (#16760)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 187, "deletions": 0}}, {"node": {"oid": "bd4d0111420508ce3819f9293282deb77cbefc03", "messageHeadline": "[C#] Rename unreleased API, add utilities (#16806)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 153, "deletions": 45}}, {"node": {"oid": "f4faceab28bd7dffbaffd737719357129eadead7", "messageHeadline": "Ignore deprecated declarations warning for TRT EP build (#16948)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 15, "deletions": 0}}, {"node": {"oid": "f8d933df31dbb8c40275fc8c150c60548f70a4ca", "messageHeadline": "[JS/Web] Register JSEP contrib ops only once per process. (#16950)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "b9d80131a72f3e3bcac86d138cc46c7b8a7e4b07", "messageHeadline": "Save optimized pre_grad graph once ready (#16816)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 22, "deletions": 14}}, {"node": {"oid": "ba49d64f670cb36760dace04cc1b68853c658332", "messageHeadline": "[WebNN EP] Support LpPool, GlobalLpPool, and Log ops (#16954)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 55, "deletions": 22}}, {"node": {"oid": "4a2a248dd7cdf81f701327b2a2eef4c9f02749fc", "messageHeadline": "remove unused comments in mac CI yml file (#16964)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "59128377913204d36fb935d0f9e66c7ec7a70fe0", "messageHeadline": "[WebNN EP] Fix bug when Pad has negative padding value. (#16878)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 41, "deletions": 15}}, {"node": {"oid": "36c5b0dcdd3b96764d799f0ddc170830d9b69f7a", "messageHeadline": "Fix onnxruntime_tvm (#16933)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "50bf310dea83dca2001e70671077bc5efd3a999f", "messageHeadline": "[CUDA] RelativePositionBias supports input with padding removed (#16923)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 254, "deletions": 37}}, {"node": {"oid": "afac67bcc3316d95752f93b9838ba3892e30f9ca", "messageHeadline": "[build] fix the CI pipeline (#16962)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 19}}, {"node": {"oid": "1fbd1ed179359361d21f335592706db4505c1d55", "messageHeadline": "[CUDA] PackedMultiHeadAttention support Bias and separated Q, K and V\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 472, "deletions": 48}}, {"node": {"oid": "e412d93b00702b5853c65ff3f4e858a008be171d", "messageHeadline": "Add lsb-release package to android custom build (#16944)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1333f73a68a21ac7ba69e03c853ff95e4b43a106", "messageHeadline": "Add ONNX 1.14 test data (#16943)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "969c95f73f9bb255c79d53dd0a385caf32dcdb1c", "messageHeadline": "[js/common] a few fixes/revises to onnxruntime-common (#16853)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 52, "deletions": 5}}, {"node": {"oid": "c4e4b98fb22b6d5d0066b65d1ebdc7a969d19f65", "messageHeadline": "replace one pool with onnxruntime-Win2022-GPU-T4 (#16953)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6046456bb6e0cae3160e84a66ef917f50752c47b", "messageHeadline": "build break: apply formatter fix (#16947)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 18}}, {"node": {"oid": "49512e558a99799d4b152af752412af101cc5d42", "messageHeadline": "[DML EP] Add I/O binding and `If` operator (#16859)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 289, "deletions": 81}}, {"node": {"oid": "ba23e5b2347b8cf212d9672784ab55e31e133ea7", "messageHeadline": "[JS/Common] Fix malformed result of Tensor.fromImage(ImageBitmap) (#1\u2026", "author": {"name": "Artyom Stepanishchev", "email": "artyom.stepanishchev@gmail.com", "user": {"login": "artemjackson"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "fa8487ea3a3a224050c816457b111c0769e0e9bf", "messageHeadline": "[js/webgpu] Check profilingMode in each run (#16897)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 10, "deletions": 7}}, {"node": {"oid": "3c72f43f7817104a756719ccf7648a99a1e32dda", "messageHeadline": "Extend saving models optimized by inference session (#16912)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 99, "deletions": 23}}, {"node": {"oid": "73ddba964f82c5b67a6142d71b83740c3cb04cae", "messageHeadline": "Update the MacOS/Linux build scripts that build/install protobuf from\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 177, "deletions": 82}}, {"node": {"oid": "28a099fca8bebab1b7014d4ff1ba53c789494d41", "messageHeadline": "unify the steps of downloading cuda sdk and setup env (#16896)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 52, "deletions": 59}}, {"node": {"oid": "50764362ac0e7fb747768593df6712fbe22c51c5", "messageHeadline": "Update protobuf Natvis visualization (#16911)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 126, "deletions": 120}}, {"node": {"oid": "77b2b618b2235dc0d5bf782584056cf3cf5b8e03", "messageHeadline": "[JS/WebGPU] Add Resize operator (#16680)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 879, "deletions": 35}}, {"node": {"oid": "3fd1d3b9bd83d557a065f50470857fd90925d6ab", "messageHeadline": "Improve graph transformer DoubleQDQPairsRemover (#16910)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 68, "deletions": 8}}, {"node": {"oid": "dd24d527375a60f64ccee9c1b378cfce91a2ab79", "messageHeadline": "[JS/Web] Added Gelu contrib operator support to JSEP (#16909)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 242, "deletions": 20}}, {"node": {"oid": "92b6e10d37c50638d59620c5a315c6e75b47131c", "messageHeadline": "skip test_smooth_quant to unblock Python Package Pipeline (#16914)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "742edec5e88786141dd5626f48adbd21d12ef4a8", "messageHeadline": "[CUDA] Add PackedMultiHeadAttention operator (#16779)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2052, "deletions": 93}}, {"node": {"oid": "7c05f7bab109cfa2ea519902920133305efccab9", "messageHeadline": "Fix IRFFT contrib op output dimension calculation (#15662)", "author": {"name": "Alexey Kamenev", "email": "alex.kamenev@gmail.com", "user": {"login": "Alexey-Kamenev"}}, "additions": 34, "deletions": 29}}, {"node": {"oid": "1743e9a615012297fc5eb58665be88dbf374927a", "messageHeadline": "[js] enable formatter for more file types (#16888)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 39601, "deletions": 103264}}, {"node": {"oid": "65534ff9ef40a2d16ed001b24e9b89ac38a838ee", "messageHeadline": "Update setup.py to add py311 (#16899)", "author": {"name": "Paul Willot", "email": "kiwi.pwillot@gmail.com", "user": {"login": "m3at"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "21a71d52bd2074b770807b209939ec11e2c64fa7", "messageHeadline": "Enable CodeQL for Android build as per 1CS requirement. (#16875)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 18, "deletions": 3}}, {"node": {"oid": "9f21f694cf939a960f5b516ba5ff9675fc3ec6d1", "messageHeadline": "stop support to VS 2019 (#16892)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 16, "deletions": 390}}, {"node": {"oid": "a021cb1b6e2035cf0a0c6841307aecc48d38227b", "messageHeadline": "Allow creating ConstantScalarNode for double type (#16797)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "634a3f2f28cf2acb357690bfda20367dbb047272", "messageHeadline": "[WebNN EP] Support Max and Min ops (#16858)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 124, "deletions": 12}}, {"node": {"oid": "9dcbcf1d2f387938c1e9478c9c2a5d55c2d7ea71", "messageHeadline": "Delete unused files (#16887)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 1043}}, {"node": {"oid": "161a9d1d6d46dd894be86383bca7fce8d36957cc", "messageHeadline": "Add some safety check for conv op (#16839)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 38, "deletions": 38}}, {"node": {"oid": "e67547b9783655dc09ab76d072edd3bd6368cfde", "messageHeadline": "[JS/WebGPU] Added Flatten operator support. (#16860)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 116, "deletions": 9}}, {"node": {"oid": "ec935a55332f8dcfabaf09e216488da614ba1292", "messageHeadline": "[QNN EP] Check the axis attribute for LayerNorm for HTP (#16872)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 14, "deletions": 3}}, {"node": {"oid": "2748f51603cfef3940866c12babd0cdd41396490", "messageHeadline": "[QNN EP] Remove duplicate string define (#16877)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 51, "deletions": 84}}, {"node": {"oid": "779fba1666ae4b53ba547666039c777bccdcf56b", "messageHeadline": "ORT Cache (#16744)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 140, "deletions": 0}}, {"node": {"oid": "bd95a8ea77296ffb88418b3fe1c6761de0229c21", "messageHeadline": "update onnxruntime-gpu-winbuild-T4 to onnxruntime-Win2022-GPU-T4 (#16\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 14, "deletions": 26}}, {"node": {"oid": "340f4ded73002cd3f192936bb2f789c156df706d", "messageHeadline": "[java] Fills out the javadoc so there are no more documentation warni\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 176, "deletions": 41}}, {"node": {"oid": "fe463d49575562d5e304b0ec79068c82ff1cf8dd", "messageHeadline": "Support SmoothQuant for ORT static quantization (#16288)", "author": {"name": "Wang, Mengni", "email": "mengni.wang@intel.com", "user": {"login": "mengniwang95"}}, "additions": 280, "deletions": 1}}, {"node": {"oid": "eeef15788804b0217d46ad6d1c2744585ad98d0c", "messageHeadline": "Format c++ code under `winml/` (#16660)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 12715, "deletions": 10781}}, {"node": {"oid": "649930142f11af6e11b11d7674883d60fda21ad7", "messageHeadline": "[DML EP] Add NCHW and float16 gamma/beta support for GroupNorm (#16814)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 587, "deletions": 121}}, {"node": {"oid": "39fca225ea90f306955d4c35b3dc46170a468d43", "messageHeadline": "ORTModule log clean up (#16795)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 393, "deletions": 181}}, {"node": {"oid": "bf006d34a9f732b6f0516e99af4229acd62d099c", "messageHeadline": "Used feature macro for if constexpr in a public header (#16836)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "d0df83e408c7a166856ed1c7b21e4a5475475949", "messageHeadline": "[WebNN EP] Support rest Reduce* ops (#16824)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 52, "deletions": 5}}, {"node": {"oid": "0c1a5098dc3417c7bf244e7b19f716164bd3aacc", "messageHeadline": "Disable PERF* rules in ruff to allow better readability (#16834)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 133, "deletions": 135}}, {"node": {"oid": "53c771f215dd30adb41e24c13619a8a4aba13802", "messageHeadline": "[js/common] add unit tests for onnxruntime-common (#16812)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1266, "deletions": 2064}}, {"node": {"oid": "03ce0a56935f512d7acfb8a4217d21187a658ad8", "messageHeadline": "[Web/JS] Added Slice operator in JSEP. (#16811)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 334, "deletions": 9}}, {"node": {"oid": "a1bb6705362a5eaf57bc7c3cfd20dd3c054df2a6", "messageHeadline": "[java] Fp16 fix for android/react native (#16832)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 577, "deletions": 284}}, {"node": {"oid": "e01365f80b640bac22b219c267b618fcebc8862c", "messageHeadline": "Update upload_pod_archive_and_update_podspec.sh to take path pattern \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 17, "deletions": 3}}, {"node": {"oid": "38db5eca65393acf4be464b9b1c3b96c78faa576", "messageHeadline": "replace onnxruntime-Win-CPU-2019 with onnxruntime-Win-CPU-2022 (#16844)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 5, "deletions": 20}}, {"node": {"oid": "feeb0b50f9e5dcf4b63b2dec3abc05a5733672f8", "messageHeadline": "Improve TreeNodeElementId hash function (#16459)", "author": {"name": "Luis Rios", "email": "lhrios@gmail.com", "user": {"login": "lhrios"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "daef133982a1c6b1f9a998e02c2172eea22f7310", "messageHeadline": "update onnxruntime_perftest's README.md as vitisai is supported on v1\u2026", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f88f0d8e36abd2d29bce4b049b232609419e9f6c", "messageHeadline": "Upgrade 4 stages in nuget pipeline to VS2022 (#16825)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 20, "deletions": 26}}, {"node": {"oid": "8b30dc11d74c070e441f961bc92b58fdde7713fa", "messageHeadline": "Update run_CIs_for_external_pr.py to skip passed checks (#16808)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 19, "deletions": 2}}, {"node": {"oid": "2e214d6e27cdac996d80cfa39a2de0364957f15e", "messageHeadline": "Workaround to upgrade VS2022 for Windows ARM build (#16826)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "f2c0470436c39f4a6fb319c7c50e110b464bb7df", "messageHeadline": "Fix slice upstream - Incompatible dimensions (#16818)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 397, "deletions": 89}}, {"node": {"oid": "b0279b14d88ed8e70372c5b336ecaa273b207133", "messageHeadline": "[DORT] Enable Dynamic Shape in DORT and Use Different InferenceSessio\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 405, "deletions": 78}}, {"node": {"oid": "4b6d9fa8518f5dd81c0887ca7ef274f847b47bf4", "messageHeadline": "Bump actions/deploy-pages from 1 to 2 (#16402)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d8d8349a1bbb6099eb2a7a1d30dff3d819363cc7", "messageHeadline": "fix: add missing nullptr of SessionOptions V2 (#16794)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "5d17bcd776f326506360187b48808c2c005dbc94", "messageHeadline": "[WebNN EP] Support Greater and Less ops (#16782)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "8ede2f139e72bcdd1092c06314e1eb2b9f37775e", "messageHeadline": "[ROCm] Optimize ROCm CI pipeline 2 (#16691)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 60, "deletions": 42}}, {"node": {"oid": "21ef14476b714c3c77a2051ea220df15a8527c5a", "messageHeadline": "Bug fix for nested control flow ops for TRT EP (#16343)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 604, "deletions": 2}}, {"node": {"oid": "40277b7f372986e3ab280ea199b6fb5a21b502e4", "messageHeadline": "Fix orttraining-linux-gpu-ci-pipeline - LargeSizeTensorUInt64Index te\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3252ff2cb7e9e29d400f4c3ed0fb123f2806e7df", "messageHeadline": "Change DML GPU pool in Windows GPU workflow use Visual Studio 2022 (#\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "b92f02ad488acac1ad1ee7bf7ba870c93fffc74a", "messageHeadline": "Bump word-wrap from 1.2.3 to 1.2.4 in /js/react_native (#16755)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "dafe11839eb740cb6a8e169286c9a5aa8f9b9a12", "messageHeadline": "Bump word-wrap from 1.2.3 to 1.2.4 in /js (#16754)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "488544b79ac7eb12672f5eca3181c9ff8d709828", "messageHeadline": "[MIGraphX EP] Fix CopyTensorAsync and add guards for stream sync Copy\u2026", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "210d29b40e3eb23b47751f3b4809616a9d825c33", "messageHeadline": "Allow --build_wasm on a mac system (#16761)", "author": {"name": "Arthur Islamov", "email": "me@daken.ru", "user": {"login": "dakenf"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "193415a1625abc8f955771c7cd8dd33650e93d4e", "messageHeadline": "[js/webgpu] reuse buffer for GpuDataManager (#16746)", "author": {"name": "Jiajia Qin", "email": "jiajia.qin@intel.com", "user": {"login": "qjia7"}}, "additions": 42, "deletions": 9}}, {"node": {"oid": "d79515041c21c42470c470c2c0617f495e626f02", "messageHeadline": "[Better Engineering] Bump ruff to 0.0.278 and fix new lint errors (#1\u2026", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 269, "deletions": 260}}, {"node": {"oid": "d3295f4329d744fe1f8419e1220e123807282b99", "messageHeadline": "[Better Engineering] Fix N802 lint errors in tests (#16788)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 174, "deletions": 181}}, {"node": {"oid": "a8e776b78bfa0d0b1fec8b34b4545d91c2a9d175", "messageHeadline": "[java] Adds support for fp16 and bf16 tensors (#16703)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 879, "deletions": 231}}, {"node": {"oid": "1e18efade5d293033900ea1c7ff1ea746fa9f17f", "messageHeadline": "[C#] Add ML Sequences and Maps Create and Process APIs (#16648)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 581, "deletions": 229}}, {"node": {"oid": "4d569f6586d109983e9059b64cd6c2dcaea643e2", "messageHeadline": "[QNN EP] Op support: LayerNorm, Asin, Sign (#16740)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 392, "deletions": 96}}, {"node": {"oid": "b508c7236f98d1fc200cec6a6df97dfd49c558d7", "messageHeadline": "Replace call to deprecated torch.norm (#16758)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "b7176f98266924053204fe25c3f2e701615c2585", "messageHeadline": "Fix bug with saving model optimized by inference session (#16716)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 79, "deletions": 13}}, {"node": {"oid": "0f9883f804e4f7847e7215950a1eabfe047543d7", "messageHeadline": "Fix Mac M1 build (#16763)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 41, "deletions": 3}}, {"node": {"oid": "538d2412ef8498b0a90fd73bb74e085844fed721", "messageHeadline": "Objective-C Add Support to Create and Query String ORTValues (#16764)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 179, "deletions": 10}}, {"node": {"oid": "a8c263f92c63e155e9c97c56ee561c8a81ec152e", "messageHeadline": "[QNN EP] Update QNN SDK to 2.12 (#16750)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 137, "deletions": 74}}, {"node": {"oid": "eaea34f8e29df9fb21fab675a3a895084407f306", "messageHeadline": "[WebNN EP] Support PRelu op (#16756)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "bb136f86c8a1d0bcbdc2a77cb16f1c26c9ebd817", "messageHeadline": "[ROCm][MIGraphX] for googletest dep, set OVERRIDE_FIND_PACKAGE (#16715)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0e40049eb277bcc468b66b46d0bf79542e424b93", "messageHeadline": "[WebNN EP] Add support for Op Pad. (#16732)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 214, "deletions": 0}}, {"node": {"oid": "2bc9fbb62141dc8ab3c9355690cf2cb2888217eb", "messageHeadline": "Fix url in the code documentation (graph optimizations) (#16770)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "c314d7724f4f2e7e95b0d94e76dbacd9f7141531", "messageHeadline": "Update dml gpu pool to onnxruntime-Win2022-GPU-dml-A10 (#16765)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "8b866060f20d9cb2d66bf027ef03813928e6666e", "messageHeadline": "Comment out ORT-Nightly feed in test app NuGet.config (#16762)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 15, "deletions": 2}}, {"node": {"oid": "fc1f463ff11e6da06d231e54bb6a10d23f50cfe5", "messageHeadline": "[ios] Enable training package in packaging pipeline (#16683)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 310, "deletions": 257}}, {"node": {"oid": "6e895fe70a227472a51e1f88f8d795234bb183d3", "messageHeadline": "Parallelize Max (#16745)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "24566058b3e5bb9e511513977cee6e7c553fd5c2", "messageHeadline": "ovep dockerfile and wheel docs changes (#16482)", "author": {"name": "saurabh", "email": "saurabh1.kale@intel.com", "user": {"login": "saurabhkale17"}}, "additions": 8, "deletions": 9}}, {"node": {"oid": "dcb0f2cdde3f686c2047334fe5985af14294ca35", "messageHeadline": "[WebNN EP] Only support opset >= 7 (#16730)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 3, "deletions": 33}}, {"node": {"oid": "7dcb805ab8377245a55d24a4aed4508ac274612d", "messageHeadline": "[js/web] upgrade onnx-proto version (#16722)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 10341, "deletions": 66}}, {"node": {"oid": "f8e3aacb477f91a404a237791c3c8942cf143a7c", "messageHeadline": "[QNN EP] Fix support of ArgMin and ArgMax on all QNN backends (#16731)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 397, "deletions": 108}}, {"node": {"oid": "95f96287762f3626f68fd09834f0b4db35da4abf", "messageHeadline": "Fix transformers optimizations for GPT-NeoX (#16743)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 105, "deletions": 29}}, {"node": {"oid": "e752cbe7f2eb27e554d7f33bf98fbbddc3c89ce1", "messageHeadline": "Work on eliminating Internal Compiler Error (#16741)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "b71ebf91a560127fa190f2ececcba7dfab5d1481", "messageHeadline": "[DORT] Reduce global configs to make enabling dynamic shape easier (#\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 77, "deletions": 68}}, {"node": {"oid": "9b549c646c87f8547e02014828bb7482a266c9ce", "messageHeadline": "[ROCm] fix kernel explorer GemmSoftmaxGemm test (#16735)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 52, "deletions": 2}}, {"node": {"oid": "9ba5cdbaa481550f86c53826c83bc094eafc2594", "messageHeadline": "[CANN EP] Fix Float16 support for CANN EP (#16733)", "author": {"name": "zhangsibo1129", "email": "134488188+zhangsibo1129@users.noreply.github.com", "user": {"login": "zhangsibo1129"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "0cab7e1a37ef8814bceb74b6285d7bbd08f20d3c", "messageHeadline": "[ROCm] Generalize FastGeLU (#16623)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 575, "deletions": 476}}, {"node": {"oid": "ad90352a6888320c539d8e4e8844e63ddca265ab", "messageHeadline": "Add MAUI test app that can be used to test model loading and performa\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 61316, "deletions": 262}}, {"node": {"oid": "a45b834722618fd5c0b08a062ff65411ba850698", "messageHeadline": "Fix warning about uninitialized member (#16736)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "df8843c4a727d1ece1c81d6334bfd63e01cb41b3", "messageHeadline": "Upgrade old Python version in packaging pipeline (#16667)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 30, "deletions": 22}}, {"node": {"oid": "b8c40b7813adbe5349b7d29dab7d1b07bfd50edd", "messageHeadline": "Fix parameter naming that fails Doc generation. (#16717)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e1ca8ee6d4a188f01efd4083d6f4273a3ef901aa", "messageHeadline": "RunAsync C/CXX API (#16613)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 308, "deletions": 66}}, {"node": {"oid": "2cf31a20cf98de3386afac7c90a37d4412d4a269", "messageHeadline": "Cuda: Decoder Masked Multihead Attention Q values get corrupted when \u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "2b7a94e65b0927cb2f7e3d4ef754600892514475", "messageHeadline": "[WebNN EP] Make some types clearer (#16705)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "2ae041f390917f776b4bb746fe9570f047c274d9", "messageHeadline": "atomicAdd returns previous value, not current value. (#16690)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "44fd98ebfee71f1d4408a228c4102030e0572f12", "messageHeadline": "[DORT] Enable aten::full by implementing extra logics to select EP (#\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 126, "deletions": 13}}, {"node": {"oid": "f236768d5c4bfd3c378400dc86fb6ee760cbaa35", "messageHeadline": "[ios] Enable `--use_extensions` with custom built iOS pod (#16711)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 29, "deletions": 0}}, {"node": {"oid": "4faee2e44c2029d7f5fac1d927f8c843b22189ca", "messageHeadline": "Fix issue in constant-propagation inside function subgraph (#16330)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 33, "deletions": 4}}, {"node": {"oid": "ea43671eb6a55702a2faa3bb3d8eee4c48cac62b", "messageHeadline": "[WebNN EP] Support several activation ops (#16693)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 110, "deletions": 67}}, {"node": {"oid": "a189e76fdef0dee1d1c110fb478cbcab05cffa95", "messageHeadline": "[QNN EP] Fix error handling for Softmax/ReduceOps (#16700)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 380, "deletions": 367}}, {"node": {"oid": "9889f0f50754671f4df397747901974865422a12", "messageHeadline": "Add support for training apis to support custom ops (#16601)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 602, "deletions": 107}}, {"node": {"oid": "19169afe30fa48db1c8b2675016e12d1461a1910", "messageHeadline": "[QNN EP] Add option to skip unit tests in the QNN NuGet packaging pip\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "853c4ff0a581e8e27bae3d3c82cf984930a0c5f7", "messageHeadline": "[C#, CPP] Introduce Float16/BFloat16 support and tests for C#, C++ (#\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 3854, "deletions": 494}}, {"node": {"oid": "77b45c6503e9fd5502d288337370faef30f6e34f", "messageHeadline": "Add Stable Diffusion Benchmark on A100-PCIE-80GB (#16702)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 14, "deletions": 3}}, {"node": {"oid": "36b121d8c2badf02eada00d59ab1ade8f761e917", "messageHeadline": "add more check to Web CI on cache restore (#16689)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 14, "deletions": 2}}, {"node": {"oid": "810512c6583fa4acf0a114627dfe7bc481f819ff", "messageHeadline": "[ROCm] TunableOp: add hipBLASLt tuning logic (#16338)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 489, "deletions": 182}}, {"node": {"oid": "a3fc04ba748e266cc98186b7974a61264736fe4d", "messageHeadline": "Fix CodeCoverage pipeline (#16684)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d1d65978f6c911cb7fb96489ba70b61c4b232da4", "messageHeadline": "[js/web] fix file size trim for wasm only .min.js (#16681)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "5de2e2fb76d90788e21271ee63692cea0c3085a1", "messageHeadline": "Call `lazy_reset_grad` in on-device training docs (#16696)", "author": {"name": "Danny Friar", "email": "dannyfriar@hotmail.co.uk", "user": {"login": "dannyfriar"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a461608409a67f7eb4f7f9badc12ca81cd20f52d", "messageHeadline": "Amx flag removal (#16527)", "author": {"name": "Dipanjan Sengupta", "email": "tukai007@gmail.com", "user": {"login": "dipanjans12"}}, "additions": 461, "deletions": 162}}, {"node": {"oid": "c07a3b869c535d15bdb89f57463483f20967f040", "messageHeadline": "Triton Codegen for ORTModule (#15831)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 6110, "deletions": 43}}, {"node": {"oid": "7cac114e522c694d669819869c965ea6f4f89751", "messageHeadline": "[WebNN EP] Support Abs and Neg ops (#16672)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "d5b76cff600b8c6375631e7263a6fac64dd8a926", "messageHeadline": "[WebNN EP] Fixed build error (#16671)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 24, "deletions": 21}}, {"node": {"oid": "b7fd5af48bcad76466b67dcd95dfc78a0bb469b2", "messageHeadline": "[ROCm] TunableOp: Update rocBLAS get_solutions API (since ROCm5.6) (#\u2026", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 320, "deletions": 394}}, {"node": {"oid": "ebc311365bf996e512b0a7ec0ea0448fa3188d63", "messageHeadline": "[ROCm] Optimize ROCm CI to reduce time (#16620)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 36, "deletions": 12}}, {"node": {"oid": "af89496fc7cab02d01d65291e1d0d29e4fc7657d", "messageHeadline": "Allow generic pipeline to accept some params for cross attention (#16\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 89, "deletions": 14}}, {"node": {"oid": "38666145196b7cf782e6761e43f5e3d9a7d54db6", "messageHeadline": "Avoid cmake repeatly printing DISABLE_FLOAT8_TYPES=ON (#16656)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "f3b40abe297706a4448b5be073e5d7772bba1d11", "messageHeadline": "Use pipeline cache to cache onnx node test data. (#16659)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 22, "deletions": 18}}, {"node": {"oid": "111382746e22d6eeb08461e5acae743c55103f15", "messageHeadline": "[js/rn] Add test for validating \"executionProvider\" options (#16651)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 110, "deletions": 27}}, {"node": {"oid": "dd7d721f3c4a31750920205b860472c342bc4eed", "messageHeadline": "support rotary embeddings in decoder masked self-attention (#16556)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1953, "deletions": 101}}, {"node": {"oid": "0c956bef0acbc315128c2151dac3943b06b61b0c", "messageHeadline": "[WinML] Fix warnings in OnnxruntimeEngine and OnnxruntimeEngineBuilde\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "2449ded20f3b48bfbb8a58d161562f61d3129647", "messageHeadline": "Use autograd_inlining for model export (#16665)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 15, "deletions": 1}}, {"node": {"oid": "596dbe277ebae209f25094631de39c31a4505b18", "messageHeadline": "[ROCm] add upgrade to fix security issue (#16668)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "ecca11340a154324e7a67c8aead05f97523b8743", "messageHeadline": "[js/common] allow creating (u)int64 tensors in 2 ways (#16541)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 48, "deletions": 23}}, {"node": {"oid": "8e393e0b8c3ec5df04604b6facfcfa1f61f25e19", "messageHeadline": "Unique operator with double (#16359)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 24, "deletions": 3}}, {"node": {"oid": "1b8d5c43c2c11340ce043d7ac9e928f366986319", "messageHeadline": "Fix builds (#16646)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 216, "deletions": 175}}, {"node": {"oid": "ce68a4c06a3364727e916a171139c5a792030378", "messageHeadline": "Fix Linux build failure when onnxruntime_DISABLE_ABSEIL=ON (#16373)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 63}}, {"node": {"oid": "2de5807703cb1553cc299ecd203cd531ab62e77b", "messageHeadline": "Attention fusion for UNet onnx model export from PyTorch 2.* (#16629)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 140, "deletions": 47}}, {"node": {"oid": "b4bf7d504428d956effaecd9abbac9a46c6205a7", "messageHeadline": "[js/web/test] accelerate 'npm test' suite0/1 init time (#16558)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 50, "deletions": 44}}, {"node": {"oid": "72076e53207f6243020ba263cb8f146cf442ab3a", "messageHeadline": "Update converter registry usage in `orttraining_test_dort_custom_ops.\u2026", "author": {"name": "Ti-Tai Wang", "email": "titaiwang@microsoft.com", "user": {"login": "titaiwangms"}}, "additions": 9, "deletions": 10}}, {"node": {"oid": "d41bbac7b9fead1ee73739d6ef4304bf08142be7", "messageHeadline": "[Web/JS] Added Expand operator support. (#16577)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 158, "deletions": 2}}, {"node": {"oid": "1b07bbceaa234a37d0418d8e61f2d52dbd80af37", "messageHeadline": "Update build.bat Prevent spaces in path (#16635)", "author": {"name": "Tommy Au", "email": "75346987+smarttommyau@users.noreply.github.com", "user": {"login": "smarttommyau"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ad994565ae66467244ba3ea1af09291de7c8988e", "messageHeadline": "Fix type annotation for InferenceSession (#16632)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "617b3a84ba5a81888cd95b11ace1feda9e96f981", "messageHeadline": "Bump semver from 5.7.1 to 5.7.2 in /js/react_native/e2e (#16653)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "3608592ef57766e07b782eeade1bb3a67ad7c9bc", "messageHeadline": "Bump semver from 5.7.1 to 5.7.2 in /js/react_native (#16652)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 11, "deletions": 18}}, {"node": {"oid": "1ebc5d38795c20412dbc049dfb312bcbbd1d6408", "messageHeadline": "Log ORTModule initialization overhead (#16529)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 142, "deletions": 31}}, {"node": {"oid": "347c963d5cdbc180980d9a7758ca6aab46da5b52", "messageHeadline": "[ROCm] Add ROCm Triton TunableOp for GroupNorm (#16196)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 451, "deletions": 167}}, {"node": {"oid": "5b6c1394cbfa74187c64ae1b6f31d9c1c4d1b331", "messageHeadline": "[js/test] CI: use pre-downloaded testdata in image (#16562)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "53057ec1f59ce4823ffb4929c6e9ceedd0e9344d", "messageHeadline": "[WebNN EP] Support InstanceNormalization and GroupNormalization. (#16\u2026", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 92, "deletions": 34}}, {"node": {"oid": "15cb2f5a8aa1bdbbb71de9560dc8835a943264d5", "messageHeadline": "Warn the user when nondet kernels are invoked in det mode (#16571)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 98, "deletions": 27}}, {"node": {"oid": "b4c4e2b594b440124436650dc41949413d85a59d", "messageHeadline": "[objc] Add session options register custom ops with function pointer \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 94, "deletions": 2}}, {"node": {"oid": "211fe5988e9329f866a64ecc3e7b9c66859d11a4", "messageHeadline": "add steps to write modulewithloss wrapper (#16486)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 104, "deletions": 0}}, {"node": {"oid": "de1b66a25a2f785cf0b482ec8a71c081dd165caf", "messageHeadline": "Bump tough-cookie from 4.0.0 to 4.1.3 in /js/react_native (#16633)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 29, "deletions": 5}}, {"node": {"oid": "b8f6235f111bd80d07939b44b6c0fd3e82fffad5", "messageHeadline": "Update stable diffusion benchmark for TensorRT EP (#16560)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2257, "deletions": 65}}, {"node": {"oid": "2fd5e1cc39e262c1232f7c0dec46548d82f8d780", "messageHeadline": "[ROCm] fix shell bug (#16641)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3b729e5d2f10ae4497816e990db17f05c89f8ca6", "messageHeadline": "[ROCm] use cupy for GPU-accelerated computing (#16611)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 52, "deletions": 33}}, {"node": {"oid": "5fee3f43024a22da84ca9b33d52208368a02d590", "messageHeadline": "Remove the special min cmake for rocm (#16570)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "2f56815344bac9a2e009774e701c2139988cfe5f", "messageHeadline": "[Whisper] Fix provider in export procress (#16545)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cb4bf4f5c8f04864fedabf11cc9bbcf13ab101a6", "messageHeadline": "[ROCm] Move ROCm build step on CPU only machine (#16596)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 204, "deletions": 108}}, {"node": {"oid": "bcebd3b1ca415e2b22b033c17a50ec0c4ecbc14c", "messageHeadline": "Allow upstream for Slice on single axis (#16410)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 238, "deletions": 0}}, {"node": {"oid": "67f4cd54fab321d83e4a75a40efeee95a6a17079", "messageHeadline": "fix JSEP build break (#16636)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "00e8f2a2a909b96240d43f73b8c9fdea097e9783", "messageHeadline": "[Web/JS] Add ConvTranspose support (#16433)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 846, "deletions": 9}}, {"node": {"oid": "5c6613875cff75b3ce41ce653707a3ad9497d2e7", "messageHeadline": "[js/web] [JSEP] allow passing data in kernel compute (#16621)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 73, "deletions": 14}}, {"node": {"oid": "329e8156d493d08900fb61f16dd9b862c657bacb", "messageHeadline": "clean unused parameter in ORT_UNUSED_PARAMETER (#16538)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 42, "deletions": 44}}, {"node": {"oid": "e55a20ece871efadef2fd005063144605abfef4f", "messageHeadline": "[Web/JS] Added Split operator support. (#16567)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 301, "deletions": 6}}, {"node": {"oid": "dd132525066b711af73fb2eeb7190998ab962502", "messageHeadline": "[QNN EP] Support 1D Conv/ConvTranspose (#16563)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1276, "deletions": 575}}, {"node": {"oid": "5933a183df3dc3896b5f7050b353d61a5f5cb01d", "messageHeadline": "[Web/JS] Added missing L1Reduce and L2Reduce oprator kernels. (#16580)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 36, "deletions": 16}}, {"node": {"oid": "01c5d05712194b4044e66cfa0520d36dad0174d8", "messageHeadline": "Avoid repeated GemmSoftmaxGemmPermuteTunableOp<HipT> ctor invocation \u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 25, "deletions": 4}}, {"node": {"oid": "47a0289ee6ff7c92e28ec67ba635cbbbaa1909ce", "messageHeadline": "[CI] Removes type2 in process_registration and fix Windows GPU Reduce\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 3, "deletions": 6}}, {"node": {"oid": "6be7b03e53ac62bb171ded8299ecde24b3686d1c", "messageHeadline": "Enable `-Wshorten-64-to-32` warning if available. (#16524)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 125, "deletions": 91}}, {"node": {"oid": "e22b0836e79c2a1c7cf422a657ce6f41d6c950b5", "messageHeadline": "[objc] Update docs and fix static analysis build (#16617)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 12, "deletions": 7}}, {"node": {"oid": "2a11f29eaa70062da22bf2845df792221952119b", "messageHeadline": "[CUDA] Optimize BiasGelu/BiasGeluGrad Kernel (#16608)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 346, "deletions": 315}}, {"node": {"oid": "697dd12f6e5f2a821656846e7587c43904fce944", "messageHeadline": "Re-organize the transpose optimization and layout transformation file\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 982, "deletions": 768}}, {"node": {"oid": "5ffd58c8e6574934a8272c7547697db32a9c8828", "messageHeadline": "Fix Reduced Ops pipeline (#16612)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "51c42ae64ae71a32e85065288e738fe585de2942", "messageHeadline": "[C#] Allow users to quickly populate native string buffers with utf8 \u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 32, "deletions": 5}}, {"node": {"oid": "d13f3153d7a1e87d81095ef5961887c1086f4b73", "messageHeadline": "[js/webgpu] enable op test for webgpu (#16542)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 309, "deletions": 22}}, {"node": {"oid": "d906d48ae9f325eaa5c89d71a4703793e4bf5001", "messageHeadline": "Support custom ops taking float 8 tensors as inputs and outputs (#16323)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 313, "deletions": 27}}, {"node": {"oid": "180292f426b5225a931f971864ad3023179529ac", "messageHeadline": "Modify CANN EP to align with the EP API refactor and fix CANN CI  (#1\u2026", "author": {"name": "zhangsibo1129", "email": "134488188+zhangsibo1129@users.noreply.github.com", "user": {"login": "zhangsibo1129"}}, "additions": 25, "deletions": 11}}, {"node": {"oid": "b84c63db2a3e822b4aac58b73714f7b871e37e9c", "messageHeadline": "Fix ORT_RETURN condition mistake (#16520)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 20, "deletions": 20}}, {"node": {"oid": "fed08e070a508bf29d11a44cf3b89ecd019f015d", "messageHeadline": "Add compiler cache in linux  wasm build (#16579)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 186, "deletions": 40}}, {"node": {"oid": "fd8ad9b950583c46c51afb02b3e4938625501d26", "messageHeadline": "Enable iOS packaging for training  (#16525)", "author": {"name": "Vrajang Parikh", "email": "vrajang@outlook.com", "user": {"login": "vraspar"}}, "additions": 139, "deletions": 60}}, {"node": {"oid": "ba914571831d7cb59583c4b64ca062b2b4e54151", "messageHeadline": "[java] Adding addExternalInitializers and addInitializer to OrtSessio\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 444, "deletions": 16}}, {"node": {"oid": "661fd4b97842308abae487e239183fdde3ff6d34", "messageHeadline": "[js/rn] always use 'typescript' from /js/ folder (#16554)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 16}}, {"node": {"oid": "a7c892106d4dfc63a6aa037d01a32dacd4bc5e0b", "messageHeadline": "[Web/JS] Support WebGPU Concat operator (#16543)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 85, "deletions": 1}}, {"node": {"oid": "d8792f80408255272a407d59d457fdcbb674a9b5", "messageHeadline": "Fix TRT EP allocator memory leak (#16552)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 17, "deletions": 3}}, {"node": {"oid": "9799d43c36ae494267f55ef7931cc546e61f2aef", "messageHeadline": "LabelEncoder kernel creation improvement (#16516)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "e2526714e23284c9a722fe9b714ef66c0e39c7f5", "messageHeadline": "[ROCm] Move MIGraphX build step on CPU only machine (#16582)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 99, "deletions": 27}}, {"node": {"oid": "a0a5f57581a74ca766c30ff72b5e49200887d00f", "messageHeadline": "[DORT] Use new FX-to-ONNX exporter (#16450)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 198, "deletions": 250}}, {"node": {"oid": "d540c7da0f47324e3548dff0dcc1329689227e01", "messageHeadline": "[ROCm] Add ROCm5.6 to python package pipeline (#16572)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "13cc6192e5d877d7c782eaaa7f9fd1544a1902f2", "messageHeadline": "[java] Adding native library loader to SessionOptions and RunOptions \u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 16, "deletions": 0}}, {"node": {"oid": "2e2e6aeff646178d5606f85c33ff4c1189dbd0ef", "messageHeadline": "[NNAPI EP] Add leakyRelu support (#16533)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 142, "deletions": 0}}, {"node": {"oid": "ac100ebb649a723d65bc73bb19051cb375a4f000", "messageHeadline": "Fix orttraining-ortmodule-distributed CI (#16569)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "f46956056d7f8ea214e794e5cbedee36f21f7ea2", "messageHeadline": "Add WinML Experimental API to Register ORT CustomOps Libraries (#16535)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 132, "deletions": 17}}, {"node": {"oid": "80a9c40cba799a6823ff67745edcc9f0e38671f7", "messageHeadline": "Bump actions/checkout from 2 to 3 (#16405)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "c8a94f1ef77fc735ffb34e408d2c5c2d9ad413dd", "messageHeadline": "Bump actions/setup-dotnet from 2 to 3 (#16403)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f5b2d213eb5c86514f5cf0c17c8bf49339772f15", "messageHeadline": "Fix nuget pipeline (#16553)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 52, "deletions": 50}}, {"node": {"oid": "47f136e2d3240e725b1c4b2bedeb5de0c2db240b", "messageHeadline": "Speed Up Whisper Export (#16504)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 19, "deletions": 9}}, {"node": {"oid": "708dec5d95876a534bd332cad478f99fe8cf6886", "messageHeadline": "[js/webgpu] allow 0 sized tensor for tensor view (#16540)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "3be6eb53c8b359703cb645ed2cb1cdf106924b7c", "messageHeadline": "[JS/Web] Fixed the output indexing in the shader code when the output\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 30, "deletions": 16}}, {"node": {"oid": "aeaa1d650f2dbc5e8b88570cb7fd5bb163680e4d", "messageHeadline": "make optimized_model_path be in temp folder instead of source model f\u2026", "author": {"name": "Mike Guo", "email": "myguo@microsoft.com", "user": {"login": "guotuofeng"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "2fd25de3600df5680eeb21b5ede9546e6731966a", "messageHeadline": "Use verbose logging in Android emulator in React Native CI (#16528)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5b3447beef68e050e5cb880426d72dbab26d4763", "messageHeadline": "Allow disable exceptions to work with ort-extensions (#16536)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "005149705598afac4b0d5b9bd8db710104cb1516", "messageHeadline": "Fix Comments (#16513)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 9, "deletions": 7}}, {"node": {"oid": "05c4566fe945dae7f0b448225e5d4eaeb4e66a6e", "messageHeadline": "[objc] Fix possible leak of OrtValue in initializer. (#16487)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 71, "deletions": 46}}, {"node": {"oid": "8fc3037ff44efad1b96f56585380fc6986c68fd3", "messageHeadline": "Support SCELossInternal/SCELossInternalGrad run with larger sized inp\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 646, "deletions": 323}}, {"node": {"oid": "322237f482c94578cabbbdd1bf853417766aa875", "messageHeadline": "[C#] Implement OrtValue APIs (#16206)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 3447, "deletions": 1461}}, {"node": {"oid": "9b2733de8e3a8df7f4796eb00964ca5b65f691bf", "messageHeadline": "[docs] Specify Objective-C max line length. (#16503)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "0c5f4924937a6d9556e99a3fb59b33f9cc7b35a6", "messageHeadline": "remove AllocatorMgr class (#16509)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 13, "deletions": 129}}, {"node": {"oid": "ff0894e540868826dd35075695ff46c04f51c1db", "messageHeadline": "Simplify gating check for CUDA Graph usage (#16491)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 54, "deletions": 61}}, {"node": {"oid": "efeb6672d6492d53882ab265d4016aafae9e2233", "messageHeadline": "Temporary optimizer support for ort format models in non minimal buil\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 24, "deletions": 15}}, {"node": {"oid": "960e320dff4fe141e6797384ee17f614b2001d14", "messageHeadline": "Objective C Training API: TrainingSession (#16374)", "author": {"name": "Vrajang Parikh", "email": "vrajang@outlook.com", "user": {"login": "vraspar"}}, "additions": 1150, "deletions": 12}}, {"node": {"oid": "6dd4e4801a563665d5f7f7dac89f91e1e9e4a61a", "messageHeadline": "Allow custom operator functions to safely propagate errors through th\u2026", "author": {"name": "Christian Bourjau", "email": "cbourjau@users.noreply.github.com", "user": {"login": "cbourjau"}}, "additions": 62, "deletions": 16}}, {"node": {"oid": "15f16ef36e8a9914f36b151e7cf8f5eaa169812f", "messageHeadline": "[ROCm] Add DecoderMaskedMultiHeadAttention (#16362)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 107, "deletions": 24}}, {"node": {"oid": "cbfbe210a8b8896b726ff11aaccdd103c58f8b40", "messageHeadline": "Fix bug that accidentally disabled training op tests (#16488)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "fb7e1f133f467a534065e2ad4ab7e6b6005bbfe9", "messageHeadline": "[Fix] TSA Upload failed in nuget pipeline. (#16476)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 45, "deletions": 4}}, {"node": {"oid": "e5270e3b4ff45f76a2038779ccfc52a0bac6ec4d", "messageHeadline": "shared allocator for on device training (#16432)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 6, "deletions": 8}}, {"node": {"oid": "1001ec93a7362bf717f368d3084846c13690e960", "messageHeadline": "Ryanunderhill/beamscorer gpu (#16272)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 838, "deletions": 286}}, {"node": {"oid": "f5e9625c369b88ce3e47d67417845276b349f557", "messageHeadline": "[QNN EP] Properly skip HTP test on x64 (#16500)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "892b1b19eab91c780f4014d21d78d629fa0cee02", "messageHeadline": "[js/rn] limit x86_64 arch in detox xcodebuild for react native e2e te\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "c3db1d362875325379d05cdc0369f8d9ffdc2240", "messageHeadline": "Replace float_t with float (#16484)", "author": {"name": "Michael Klimenko", "email": "mklimenko29@gmail.com", "user": {"login": "MKlimenko"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "4768ac5f300be9e5e47ceee8c3e647330cc0989c", "messageHeadline": "Fix onnxruntime-CI-nightly-ort-pipeline Failure (#16495)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a49bb85cfec6f0c4da0a483aaa235a4841b9aedf", "messageHeadline": "Manage ORTModule configurations consistently (#16396)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 506, "deletions": 407}}, {"node": {"oid": "403bebfb5102ad6aeab8d1bed873af53c3f9723b", "messageHeadline": "Use PadAndUnflatten to replace GatherGrad for restore (#16429)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 491, "deletions": 199}}, {"node": {"oid": "ae6da03438970f25ea7a555a20522faca18762ee", "messageHeadline": "ROCm decoding (#16339)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 193, "deletions": 124}}, {"node": {"oid": "6e9541046e4e1e54a2d6f367657068bc8724acdf", "messageHeadline": "extend react native ci timeout limit (#16469)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "eb4e6d206233a7a62c3f238cf2edea2b37684719", "messageHeadline": "Support Mul and Sub in padding elimination (#16478)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 59, "deletions": 37}}, {"node": {"oid": "4a331ef66726dbf27221b4ac436a7919704b15d2", "messageHeadline": "Rework CPU MeanVarianceNormalization kernel to support arbitrary axes\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 527, "deletions": 218}}, {"node": {"oid": "e2c214d81f8394156cd96ce245118b5480a9cdf0", "messageHeadline": "[TensorRT EP] TRT 8.6 minor version update (#16475)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 24, "deletions": 24}}, {"node": {"oid": "1f60414bc26e7c868c661aa96ab75498e2aff70e", "messageHeadline": "Load CheckpointState from a buffer (#16457)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 243, "deletions": 27}}, {"node": {"oid": "efe0af37206ad8a0bf8617e4a9d45adf671fea98", "messageHeadline": "[TensorRT EP] Fix nullptr check (#16468)", "author": {"name": "Yifan Li", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "c55c6255e0565b3270f0dcbbb8ef58611a1c4228", "messageHeadline": "Eliminate safe nodes that are followed by a shape node. (#16065)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 199, "deletions": 0}}, {"node": {"oid": "48eff096648415f0250a48cd2a7c73b2b550762d", "messageHeadline": "Fix file list for test of build with IO debug (#16474)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "7e211f0e034a1746cbc40ed8cbad5aad71f13569", "messageHeadline": "[ROCm] Move mount data step into docker container (#16471)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 10, "deletions": 32}}, {"node": {"oid": "8971af72af402ee6931729d24b82a10589798c1f", "messageHeadline": "fix webnn build (#16464)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "9206b7cdc61df811ace631c25428fdf7f1a1b687", "messageHeadline": "Zhijxu/cast propagation softmax (#16408)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 111, "deletions": 27}}, {"node": {"oid": "9407c3270c10352ad7e55a9bcb16f733afb23c79", "messageHeadline": "GPT-2 attention fusion for transformers >= 4.27 (#16461)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 17, "deletions": 13}}, {"node": {"oid": "a8c313dec41150e8930a6154ad9ced687958f543", "messageHeadline": "[QNN EP] Python script to modify Onnx model to make it aligned with c\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 296, "deletions": 0}}, {"node": {"oid": "d5b606d50d5a251910a45cfcef001cb7c314c8ea", "messageHeadline": "Remove now duplicated symbol (#16458)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "5c125b4366071cee21eb93945824adbaf96468d6", "messageHeadline": "Cfu revertamx (#16455)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 162, "deletions": 475}}, {"node": {"oid": "04dbdc96bf47518a7058f18ff64b027759071995", "messageHeadline": "[js/rn] Fix React Native CI pipeline E2E test (#16447)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 11, "deletions": 789}}, {"node": {"oid": "10ba1e270c522107eb32652f7cec0feed5814a44", "messageHeadline": "Minimal Build for On-Device Training (#16326)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2232, "deletions": 755}}, {"node": {"oid": "97f4484df9ee527e99d3730b98df41c90b7f419f", "messageHeadline": "Bump actions/setup-python from 3 to 4 (#16404)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "8e8840f1de86f30304800126224d48c0ed1d2e51", "messageHeadline": "Enable Web CI on Linux (#16419)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 372, "deletions": 23}}, {"node": {"oid": "a270d8407efb0a3e67724b6581b05a4a319bc8a3", "messageHeadline": "Allow saving of large models after optimization (github issue 12882) \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 50, "deletions": 2}}, {"node": {"oid": "89f8f20a61ceac0276c42d894c38400d86b175d0", "messageHeadline": "fix protobuf copyfrom 2G limit (#16422)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 123, "deletions": 154}}, {"node": {"oid": "4b69226fca914753844a3291818ce23ac2f00d8c", "messageHeadline": "Fix input typo in Whisper export with beam search (#16439)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 46, "deletions": 13}}, {"node": {"oid": "42489a8a241e0cefc0efaaf07d192b7f840cc5f9", "messageHeadline": "Add ability to create ort format models from training offline utility\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 92, "deletions": 30}}, {"node": {"oid": "0ad0d6ebbfde87d4e6b3ba34d23a832570dd34e7", "messageHeadline": "Unblock Linux MultiGPU TensorRT CI (#16446)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "de476c8075a5db2b47aea76cce9c92b2166a7cce", "messageHeadline": "[js/web] update webgl context creating (#16436)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 21, "deletions": 7}}, {"node": {"oid": "3c2d52a995954cd9ce24b2831060c4326ab18e1c", "messageHeadline": "DORT support for custom ops (#16392)", "author": {"name": "pallavides", "email": "pallavides@users.noreply.github.com", "user": {"login": "pallavides"}}, "additions": 124, "deletions": 19}}, {"node": {"oid": "da532f3f5a0e2c5a965c923ea3e5f2aa2ffb807b", "messageHeadline": "[js/webgpu] fix GPU to GPU memcpy (#16393)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "52e2bdf5417ac52ce61fbef4ae6e12fe3e1ea84d", "messageHeadline": "Add license header to CUDA related files (#16437)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 40, "deletions": 13}}, {"node": {"oid": "6e29e185f31883f9820ad69d92d83ceada7a3288", "messageHeadline": "Clean AzureEP logics  (#16367)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 3, "deletions": 868}}, {"node": {"oid": "4e3cff60fdd1e84e368de536975ea34278501b90", "messageHeadline": "CUDA graph support for TRT EP (#16081)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 282, "deletions": 85}}, {"node": {"oid": "961fa7274ad1631bb5364ef6f836d99558b9d3b0", "messageHeadline": "[NNAPI doc] add reducemean to supported op list (#16414)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b4b126ffb0abd3d0d40e0946b541d4db1b495cd7", "messageHeadline": "Set onnxruntime-c local pod path environment variable for react nativ\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "9a80955c4592e8629dd8f623e0a4d1bca7c80b33", "messageHeadline": "Add compute capacity to trtep engine cache file (#16356)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 28, "deletions": 10}}, {"node": {"oid": "64b22cd00f3af5c07552188773f4ec159f68044b", "messageHeadline": "[WebNN EP] Support Where Op (#16380)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 67, "deletions": 0}}, {"node": {"oid": "e2381c42f25caffc6c3ea524e3af561f1baa9d73", "messageHeadline": "Use M_PI to replace 3.14 constants (#16421)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "48e6186b1a41ec99816588e7235c57bdc15c4162", "messageHeadline": "Move tests from core/providers/cuda/test/* to test/providers/cuda/ an\u2026", "author": {"name": "Yuhong Guo", "email": "guoyuhong1985@outlook.com", "user": {"login": "guoyuhong"}}, "additions": 547, "deletions": 409}}, {"node": {"oid": "e83993bbafce60f3698a43933bb01ddb03610636", "messageHeadline": "Added MatMul tests for QNN EP (#15956)", "author": {"name": "Aung T Naing", "email": "aung.thetnaing@gmail.com", "user": {"login": "aungthetnaing"}}, "additions": 180, "deletions": 0}}, {"node": {"oid": "7b51a1b17d0952089c77afe705d55fa748f0aa23", "messageHeadline": "Enable Microsoft.AI.MachineLearning NuGet with WinUI projects (#16415)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c8de3eaac68010c719bb24114f26e6467ead1b3e", "messageHeadline": "Update DORT to follow PyTorch changes (#16394)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 9, "deletions": 18}}, {"node": {"oid": "30bb0959dc55b823bad9a7d774a8fca85c9358e6", "messageHeadline": "[NNAPI EP] Add ReduceMean Op support (#16294)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 232, "deletions": 9}}, {"node": {"oid": "d190db7fcdf6021a69580c2c9ae0fd2ef4eb646b", "messageHeadline": "Update default external_data_location for pre-process of quantization\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b8917ad84f9b96d664bc7062d42f188c1e60452f", "messageHeadline": "[js/web] fix nodejs detection (#16400)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 19}}, {"node": {"oid": "12dffef768adb42d4ba5b6c9347231ac1f930a9f", "messageHeadline": "added support for cmake \"find_package\" (#8919)", "author": {"name": "Prateek Chokse", "email": "prateek.chokse@gmail.com", "user": {"login": "prateek9623"}}, "additions": 82, "deletions": 39}}, {"node": {"oid": "dd72192cf4866e7ca036a7987a5ef76f930ded3a", "messageHeadline": "ExecutionProvider API refactor - move allocator from EP level to Sess\u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 889, "deletions": 1601}}, {"node": {"oid": "5dcaf70501e6346ca1e207a9aceee4ec07663f8a", "messageHeadline": "Adding this set_to_none flag to zero_grad to have signature parity wi\u2026", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "470d6c1ccee975741573e7ee744e6cbea65da6bf", "messageHeadline": "[ROCm] Delete unused file to fix Component Governance Alert (#16407)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "341484e67c8c8957f02a82d4023ddcdacd890cde", "messageHeadline": "Embedding sparsity optimization (#16141)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 1158, "deletions": 65}}, {"node": {"oid": "1418d8728cb91b21d7baa3d619aeaf5a2df41149", "messageHeadline": "[ROCm] Fix CI Pipeline (#16409)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "8b9eab093b6e98e1a08f52f660ed4b1776f68612", "messageHeadline": "keep symlinks in maven package (#16376)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 26, "deletions": 6}}, {"node": {"oid": "35fa6af428cd518ae52453c5eeef862c69cd1c02", "messageHeadline": "Fix for the build break in AMX feature on Mac OS. (#16390)", "author": {"name": "Dipanjan Sengupta", "email": "dipanjan.sengupta@intel.com", "user": {"login": "dipanjans12"}}, "additions": 15, "deletions": 9}}, {"node": {"oid": "8fdfd20191a89b1c9cd77b303701ef90a69924d8", "messageHeadline": "Separate out operator vs model testing. (#16228)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3166, "deletions": 3096}}, {"node": {"oid": "a6ce7b339f1ddf0e36a1d93927d66735f24b1fb9", "messageHeadline": "Enable model subgraph execution in OVEP and setting the OpenVINO dll'\u2026", "author": {"name": "saurabh", "email": "saurabh1.kale@intel.com", "user": {"login": "saurabhkale17"}}, "additions": 389, "deletions": 343}}, {"node": {"oid": "3f7f90aed02a0d8d99c48fa89201759477794b8d", "messageHeadline": "Stabilize Whisper export with beam search (#16297)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 116, "deletions": 36}}, {"node": {"oid": "dd660c054ea1263a3a235c5fdc3c218fb9776397", "messageHeadline": "Bump transformers from 4.24.0 to 4.30.0 in /tools/ci_build (#16331)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d813d991b17dfba73051c2c69ec20c0d46b03190", "messageHeadline": "[WebNN EP] Support Squeeze Op (#16361)", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 60, "deletions": 34}}, {"node": {"oid": "fbf08c4b4dce5da245189203d9f6cfc41f6663a2", "messageHeadline": "Fix minor TRT EP provider option issue (#16107)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "4915191e63ee014489c331053677e0f2e069008d", "messageHeadline": "Fix build of Python wheel on Windows with single-config generator (#1\u2026", "author": {"name": "Silvio Traversaro", "email": "silvio@traversaro.it", "user": {"login": "traversaro"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "685816bb0a648e4467b8fa710a301ef3bd3d5774", "messageHeadline": "[js/rn] Add executionProviders support (#16233)", "author": {"name": "Jhen-Jie Hong", "email": "developer@jhen.me", "user": {"login": "jhen0409"}}, "additions": 104, "deletions": 0}}, {"node": {"oid": "ea1a5cf92045ee3ae4af85ab71844eb0aa67e379", "messageHeadline": "[js/rn] Implement blob exchange by JSI instead of use base64 (#16094)", "author": {"name": "Jhen-Jie Hong", "email": "developer@jhen.me", "user": {"login": "jhen0409"}}, "additions": 935, "deletions": 141}}, {"node": {"oid": "9110e5b9bde306a409e86e71db3bdc5a2dda3829", "messageHeadline": "[ROCm] Add attention kv cache for decoding (#16076)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 435, "deletions": 88}}, {"node": {"oid": "96471491d705559aac6a329187fec0c744329486", "messageHeadline": "Fix test failure in debug CUDA build (#16370)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 11, "deletions": 6}}, {"node": {"oid": "1866a9d81879dd4294b4c8e105f916a4739664bc", "messageHeadline": "Use the lowest float for causal mask (#16369)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 110, "deletions": 137}}, {"node": {"oid": "bcdb81c56308e9fe4cfe0796cf71d949318ca8df", "messageHeadline": "[Whisper] add a fusion option to split input bias from MHA/DMHA (#16049)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 177, "deletions": 36}}, {"node": {"oid": "6949cfaf94f97e93463d77b304a3a5cdb19aa1de", "messageHeadline": "Fix MS domain QuantizeLinear and DequantizeLinear type registrations \u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "188d5f5398936d35649c2a6cdcaa76b3735c1653", "messageHeadline": "Fix Linux Multi GPU build pipeline (#16368)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 33, "deletions": 70}}, {"node": {"oid": "5754cd7d1d0fa0a972313fff8e039c9673c3dc8b", "messageHeadline": "Add fp16 support to CPU EP gemm op (#15506)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 671, "deletions": 536}}, {"node": {"oid": "67093b204dc449527c9d90cbd95a79711a5ab7d6", "messageHeadline": "Clean up aarch64 quantized GEMM dispatch (#16120)", "author": {"name": "Skand Hurkat", "email": "skandhurkat@users.noreply.github.com", "user": {"login": "skandhurkat"}}, "additions": 11, "deletions": 9}}, {"node": {"oid": "5c0d5768e783c43a2ec820a83394b8c680145d5f", "messageHeadline": "make package.json more rebost (#16366)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "63f557335469f07e80e7868f7b8946d41ad0d842", "messageHeadline": "Relax node placement check for CUDA Graph usage (#16358)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 100, "deletions": 13}}, {"node": {"oid": "681a0d084df0810c7e518a6deadf66265d30868a", "messageHeadline": "Removing AMX build flag  (#16086)", "author": {"name": "Dipanjan Sengupta", "email": "tukai007@gmail.com", "user": {"login": "dipanjans12"}}, "additions": 472, "deletions": 165}}, {"node": {"oid": "65434dce57ee014e5ac5dc6c734bc926ee892498", "messageHeadline": "Bump decode-uri-component from 0.2.0 to 0.2.2 in /js/react_native/e2e\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "4f7900b55365ebb2641e45f557cbf02bff4590fc", "messageHeadline": "[js/web] enable ONNX Runtime Web error messages in JS (#16335)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 283, "deletions": 207}}, {"node": {"oid": "3e99e43a1d861f7c53ee2e7e1e7d9e213c122c85", "messageHeadline": "extend Final AAR testing timeout limit (#16340)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "735a32fee11e8a5ddb269afe0cecb350cbee28ca", "messageHeadline": "Introduce memory observer for ORTModule (#16213)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 163, "deletions": 20}}, {"node": {"oid": "574e17ade41e23f52d24f0c2b6640c006f3bfdcb", "messageHeadline": "Fix Reshape check (#16349)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 75, "deletions": 9}}, {"node": {"oid": "097346be9d455af1800c2817e027d69092a5b983", "messageHeadline": "[ROCm] Add clean step for ROCm CI pipeline (#16336)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "5eec24837ffa4d0519c242896f46a02d25c8e412", "messageHeadline": "Fix for AMD GPU pipeline (#16357)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "73dad4452bec19864312bd1e56c45c7d0d8f7019", "messageHeadline": "[WebNN EP] Support Shape op (#16282)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 99, "deletions": 0}}, {"node": {"oid": "dbc7a195b17d538c64ea0757f45749db0edc4004", "messageHeadline": "Update win-ci-pipeline.yml: enable xnnpack tests (#16244)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 130, "deletions": 71}}, {"node": {"oid": "9be133231fe7585ed6e3337389058bd5ca8db89e", "messageHeadline": "Fix cuda graph capture (#15005)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 198, "deletions": 79}}, {"node": {"oid": "8a3de16d148a3ec943f9c192614bc9f727ad7729", "messageHeadline": "Temporary fix to make the training pipeline green (#16353)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ed2482667b1d99f5c1cb6ad9fd136e291f1728be", "messageHeadline": "Fix training pipeline (#16342)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c5176ed1225f03540009dc16c0e70856454152ad", "messageHeadline": "[WebNN EP] Add several new unary Ops (Ceil, Exp, Identity, Reciprocal\u2026", "author": {"name": "zesongw", "email": "zesong.wang@intel.com", "user": {"login": "zesongw"}}, "additions": 27, "deletions": 1}}, {"node": {"oid": "4f23577cb52334d84e7bbf7fb3404b4ceda2531b", "messageHeadline": "[React Native] Publish E2E test logs on build failure too. (#16327)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "e3e4926d008fb437e7a3ce40371caebe1ea0911b", "messageHeadline": "[js/common] allow import onnxruntime-common as ESM and CJS (#15772)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 222, "deletions": 127}}, {"node": {"oid": "0df9e429605c94fee79c97f1a0ed0573de15ac1b", "messageHeadline": "User/sheilk/register div nonzero (#16309)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "889f80082f9ae5581c2e9c2a00e2c27c06d5e341", "messageHeadline": "[js/web] Added Reduce operators support (#16122)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 531, "deletions": 65}}, {"node": {"oid": "40bcc0441bf1b39c320d6fc30d8be503c7a218f4", "messageHeadline": "Enhance StatisticsSubscriber (#16098)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 245, "deletions": 46}}, {"node": {"oid": "eed02a3f782407e569c29a8a86c58a4d398d0b0e", "messageHeadline": "Xnnpack QDQ test (#16281)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 3, "deletions": 10}}, {"node": {"oid": "97751ad5160fefa7d3e80382150aba23fe6024d1", "messageHeadline": "[CANN]  Fix registration of Identity operator (#16210)", "author": {"name": "zhangsibo1129", "email": "134488188+zhangsibo1129@users.noreply.github.com", "user": {"login": "zhangsibo1129"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "5ab51694ab17b93688da6533c6be0b80fc831933", "messageHeadline": "gather OP with scalar indice in NNAPI EP (#16279)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 94, "deletions": 3}}, {"node": {"oid": "59f42cccb88eb62d82e3ab504c6e6c61f532d966", "messageHeadline": "[js/common] refactor tensor type in onnxruntime-common (#15843)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 716, "deletions": 621}}, {"node": {"oid": "f274bbb0c8121f47b1101327edbe9b145b0fcda8", "messageHeadline": "[js] add API that allows to get package version (#16207)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 135, "deletions": 9}}, {"node": {"oid": "3b5a8352c1e221ad3a203c05018a0d57d7a2fc12", "messageHeadline": "CodeSign Mac packages in nuget pipeline (#16291)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 145, "deletions": 60}}, {"node": {"oid": "1a22d245e2c3d686168dc51a37aa99f6f0d9bed5", "messageHeadline": "[QNN EP] Fix auto_pad handling for Conv operator (#16299)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 79, "deletions": 8}}, {"node": {"oid": "b668a6da9622a0ba91bed06affc8d4c468bc5496", "messageHeadline": "Treat Objective-C static analysis warnings as errors (#16293)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "443f553782680ed7efff6f779eacaba404e1b3fd", "messageHeadline": "Fix native onnxruntime library not loading in Azure App Service (#16286)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 14, "deletions": 6}}, {"node": {"oid": "a9d47f72a4aca180e652bf3a0e1f9e7e75ebfebc", "messageHeadline": "[QNN EP] Add model description into context binary file metadata for \u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 88, "deletions": 61}}, {"node": {"oid": "d1e8d4a2613f25823d997a38d51b138ca9adcedb", "messageHeadline": "[QNN EP] Fix an issue for Conv with dynamic weights (#16235)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 84, "deletions": 10}}, {"node": {"oid": "ac8444f299d3f2a8eb9fdf9a921a8d9f2c651be1", "messageHeadline": "[js/rn] Implement dispose native method (#16131)", "author": {"name": "Jhen-Jie Hong", "email": "iainst0409@gmail.com", "user": {"login": "jhen0409"}}, "additions": 105, "deletions": 7}}, {"node": {"oid": "b48628f1cd519b7662d28b3916160efaacc1184b", "messageHeadline": "[QNN EP] Add tests for large inputs that trigger memory alloc errors \u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 226, "deletions": 0}}, {"node": {"oid": "b72fe664c14b55e62149105291fe6639f02da282", "messageHeadline": "Refactor prepack buffer code (#16280)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 89, "deletions": 80}}, {"node": {"oid": "9d52632da93f45454044a64de4d5bd92b9929e7b", "messageHeadline": "[DML EP] Register Div with int64 and NonZero with bool (#16276)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 11, "deletions": 10}}, {"node": {"oid": "79e02300025e63beb5e27ca00afa2773665f6c0f", "messageHeadline": "Add vocab masks to Whisper export with beam search (#16180)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 67, "deletions": 15}}, {"node": {"oid": "a3a443c80431c390cbf8855e9c7b2a95d413cd54", "messageHeadline": "Support re2 == 2023-06-02 (#16257)", "author": {"name": "Yuriy Chernyshov", "email": "thegeorg@yandex-team.com", "user": {"login": "georgthegreat"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b07b647f66b1b02514aae41e0710a698b73fed9e", "messageHeadline": "Fix some issues with NNAPI Softmax (#16095)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 156, "deletions": 49}}, {"node": {"oid": "dc1312cfb150e4eaf138b30e5c7fc85a6346a246", "messageHeadline": "[web] fix: Provide typings for exports (#16249)", "author": {"name": "Artur", "email": "artur@vaadin.com", "user": {"login": "Artur-"}}, "additions": 12, "deletions": 3}}, {"node": {"oid": "fe0cc8ce6275c4db6c7bd5945dc5b5ba16a95335", "messageHeadline": "Remove some usages of CUDA_VERSION macro (#16199)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 29, "deletions": 204}}, {"node": {"oid": "908e94066006734828e449ebd16371029ff814d1", "messageHeadline": "[CPP Api] Remove deprecated CustomOp API (#16256)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 403}}, {"node": {"oid": "67f4a4fd167f52670cf3a401116a5b153cd34b37", "messageHeadline": "Objective-C binding for ORT training  (#16127)", "author": {"name": "Vrajang Parikh", "email": "vrajang@outlook.com", "user": {"login": "vraspar"}}, "additions": 400, "deletions": 15}}, {"node": {"oid": "bca49d62a0f4775a71f914dc339aebb62b8a88d0", "messageHeadline": "Fixing CoreML in Java (#16231)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "1261d0b8ba5620579cd4d92e456a09fb8a9b369f", "messageHeadline": "Fix some build issues on MacOS with Xcode 14.3. (#15878)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 267, "deletions": 130}}, {"node": {"oid": "b8858f034e58a575c451e3fce6250aef2d80c4fd", "messageHeadline": "[QNN EP] Increase conv test tolerance for Windows x64 (#16241)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 24, "deletions": 4}}, {"node": {"oid": "a8c2f24ae0e859e28aa0c5968f042c536fa21187", "messageHeadline": "[WebNN EP] Merge support for segment anything into main branch (#16208)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 2121, "deletions": 178}}, {"node": {"oid": "05bea0d3c3ff7180c306223088916eee7aabe3ad", "messageHeadline": "Add new cases for non biased mha tests (#16097)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 323, "deletions": 10}}, {"node": {"oid": "33731608639476f2cf3d3555975e13f8549b8554", "messageHeadline": "[CPU EP] Refactor CPU mha (#16247)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 34, "deletions": 66}}, {"node": {"oid": "f0139658316ce231aa5db54b3bc575cd85c6dd4c", "messageHeadline": "Add non qkv biased version mha unittests  (#16075)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 238, "deletions": 48}}, {"node": {"oid": "3c2a11f2f1d8aeaf30f47402a5598dbd30990e9c", "messageHeadline": "[java] Allow the creation of boolean tensors from ByteBuffer (#15556)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 45, "deletions": 5}}, {"node": {"oid": "a95f8ae53cf3b6eb86c7f83361bba527fed1e870", "messageHeadline": "[ROCm] Update ROCm/MIGraphX CI pipeline (#16215)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 183, "deletions": 322}}, {"node": {"oid": "18c97381cd69db860f698451ef87722466b3542e", "messageHeadline": "Detect fake tensor mode if it has already been created.  (#16220)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "2e66bc8669d56d7730a9da0a93e97711867db03c", "messageHeadline": "prevent object destruction compile error (#16134)", "author": {"name": "Somdev Sangwan", "email": "s0md3v@gmail.com", "user": {"login": "s0md3v"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "6b5b79872be6bc4107e796bef3505c491a1876c3", "messageHeadline": "Avoid taking dependency on dl.fedoraproject.org (#16202)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 25, "deletions": 28}}, {"node": {"oid": "7686193c401f568e06588a28a69ab9f2d61ad968", "messageHeadline": "Fix DNNL build (#16201)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "319a0dc6aa98023193817a2478d99d472d3a5c24", "messageHeadline": "[js/doc] allow deduplicate opset version (#16182)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "6e1c3003ffc5a72d06c95ee381694e198ee0a2fb", "messageHeadline": "DML EP and MLAS buffer allocator - increase alignment to 64 bytes for\u2026", "author": {"name": "Dale Phurrough", "email": "dale@hidale.com", "user": {"login": "diablodale"}}, "additions": 3, "deletions": 8}}, {"node": {"oid": "5a4c3b7937c34cad9671e6440a03aa024a98bf19", "messageHeadline": "[QNN EP] Support Equal, Less, LessOrGreater, Greater, GreaterOrEqual \u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 281, "deletions": 12}}, {"node": {"oid": "f72dc198c66e8861e6015e44d36e727d0361200b", "messageHeadline": "[QNN EP]Add UT for cached Qnn context binary (#16184)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 45, "deletions": 5}}, {"node": {"oid": "5bfa1183d1a9fd8d1c8fbd81667dfb88052330cc", "messageHeadline": "Add a Memory Profiling build job in post merge pipeline (#16172)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 28, "deletions": 39}}, {"node": {"oid": "e6c6184fee5afbfa42190e29677068035db70963", "messageHeadline": "[JS/WebGPU] Unsqueeze operator implementation (#16138)", "author": {"name": "Alexander Visheratin", "email": "visheratin@users.noreply.github.com", "user": {"login": "visheratin"}}, "additions": 148, "deletions": 14}}, {"node": {"oid": "5b0817631495c9205e3bc69c627ce0b320a68abd", "messageHeadline": "Exclude shufflenet from DNNL's model tests (#16126)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d185bf444de1dd5b573da8d089d4a97417cf68d4", "messageHeadline": "[CANN] Add IOBinding Support For CANN EP (#15802)", "author": {"name": "FFFrog", "email": "85333600+FFFrog@users.noreply.github.com", "user": {"login": "FFFrog"}}, "additions": 150, "deletions": 22}}, {"node": {"oid": "8c85d990c237aed8fd0fb0931ccf812435881c4a", "messageHeadline": "add third-party pipeline status to README.md (#16155)", "author": {"name": "FFFrog", "email": "85333600+FFFrog@users.noreply.github.com", "user": {"login": "FFFrog"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "1b518c6836b8fa0ce930e8ee2eb2da4b5c9e2b52", "messageHeadline": "[ROCm] add early stop to tunable profile progress (#15716)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 96, "deletions": 6}}, {"node": {"oid": "65b316a138e93f2096009b6ada15a4cf98c4e47d", "messageHeadline": "Consolidate ORTModule logging (#16078)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 401, "deletions": 296}}, {"node": {"oid": "d19e5c0abbd2505f7e5a859be1e28366127749b4", "messageHeadline": "Fix a misaligned error in CUDA GEMM (#16130)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 26, "deletions": 9}}, {"node": {"oid": "f67f7c0f0b8e654a0b2da6aa3a57d2b65827acb3", "messageHeadline": "[js/web] disable node fallback in webpack (#16166)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "13d6ac74de66dfb38c049f9a3dfb08e9f0c80751", "messageHeadline": "fix memory profile build (#16177)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 16, "deletions": 16}}, {"node": {"oid": "a55637a103e0d4aa89c4926e7105c2404e6ac46c", "messageHeadline": "Bump socket.io-parser from 4.2.2 to 4.2.3 in /onnxruntime/test/wasm (\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "3cca32beec53d2b176a5b2846e36a74eb40a8bd9", "messageHeadline": "[QNN EP] exapand convolution test coverage. (#15975)", "author": {"name": "Aung T Naing", "email": "aungnaing@microsoft.com", "user": {"login": "aungthetnaing"}}, "additions": 72, "deletions": 12}}, {"node": {"oid": "e0199cfbd907e74396a5ebaaf66aa293fa512012", "messageHeadline": "extend mac packaging timeout limit (#16173)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ba5f5e319807126eb35ced2cfabd56a7e358ad15", "messageHeadline": "[js] allow manually release inference session (#16169)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 13, "deletions": 0}}, {"node": {"oid": "3dc5179a36b7c33471b38615c98e9136d7f66c78", "messageHeadline": "[ROCm] Change ortmodule test (#15884)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "03216e23136560e71658105343e076c021d466d2", "messageHeadline": "Bump socket.io-parser from 4.2.2 to 4.2.3 in /js/web (#16068)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "7edc4b105dae783207d7494734e4c4a00ef6d4f4", "messageHeadline": "Copy missing training header files to the package archive (#16119)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "28026148467dcbfca05a87e51535db9b92532f10", "messageHeadline": "Condition the usage of variadic callback by version (#16112)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "ebe715a8172b132c3ae577b79604c04b45a3dccb", "messageHeadline": "[js/webgpu] fix RangeError in buffer download (#16165)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bf05d4ec267832e4609446d59092b6db555764c4", "messageHeadline": "Fix nightly ort CI pipeline (#16162)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e726151b5cc1986e5b4b7fb178ecb538b6e81f67", "messageHeadline": "Introduce float 8 types (#14731)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 6706, "deletions": 1608}}, {"node": {"oid": "abd94b65b79bbca72e0640664018bd629683dde3", "messageHeadline": "eigen.cmake use url info from deps.txt (#16129)", "author": {"name": "\u795e\u697d\u5742\u5e15\u742a", "email": "mudream4869@gmail.com", "user": {"login": "mudream4869"}}, "additions": 9, "deletions": 7}}, {"node": {"oid": "90e8c8daaf96efe64d2ed60b311f0ee96a4ff34c", "messageHeadline": "profile_explorer: add op-kernel correlation info (#15946)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 186, "deletions": 45}}, {"node": {"oid": "31fc25d2c223f6ba4e0c9a297b6f91767705a398", "messageHeadline": "[Fix] Check if CUDA is downloaded in AGENT_TEMPDIRECTORY (#16142)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "6abdc3a87b290278cfb843fa375c10071bd6c98e", "messageHeadline": "Fix static analysis bug (#16114)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 26, "deletions": 25}}, {"node": {"oid": "73584f9360dbf8ba149f2c291f0bb190bf662ae2", "messageHeadline": "More fixes on nuget pipeline (#16091)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 14, "deletions": 10}}, {"node": {"oid": "415c26e46e8fdd1ccc4a39b934de1f9e17bcaa30", "messageHeadline": "[JS/WebGPU] Squeeze operator implementation (#16024)", "author": {"name": "Alexander Visheratin", "email": "visheratin@users.noreply.github.com", "user": {"login": "visheratin"}}, "additions": 103, "deletions": 0}}, {"node": {"oid": "5e41d1600a8e8209b84f24a0f12de9af29944658", "messageHeadline": "Add new QNN CIs to azp run tool (#16109)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "9939092e71dd9b6be375a86b48bc37a7a7891645", "messageHeadline": "[CPP API]Fix constness in C++API (#16103)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 43, "deletions": 8}}, {"node": {"oid": "54fdb640fe072384d2d05296c9d0711713e0e8f3", "messageHeadline": "Address performance regression with duplicate initializers across DML\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 21, "deletions": 7}}, {"node": {"oid": "a5410515ad9d39571fd70b7fd9f4c30634527c25", "messageHeadline": "Fix: Some fields in OrtCUDAProviderOptionsV2 struct are not initializ\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 17, "deletions": 29}}, {"node": {"oid": "4ab7d410aeacc4bd390d47a86384896d7d2aa599", "messageHeadline": "ExecutionProvider API refactor - Deattach allocator from EP by creati\u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 25, "deletions": 66}}, {"node": {"oid": "4bfb8d3303c72727f9fbfa529abdacd7b2fed0e5", "messageHeadline": "Update calls to OrtArenaCfg constructor to pass additional parameter.\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "2cf0ae7d0111e08e1a471bec0259d91bad205e4e", "messageHeadline": "[ROCm] Add AttentionMode to make attention logic streamline (#15978)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 349, "deletions": 79}}, {"node": {"oid": "b28e927ca4ee6ef40818bc46a82a6b9346dcb27b", "messageHeadline": "Read AA64ISAR0_EL1 to check dot product support (#16082)", "author": {"name": "Skand Hurkat", "email": "skandhurkat@users.noreply.github.com", "user": {"login": "skandhurkat"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "0d1a8cc65134de0ed7abf39f4358415b6ad35a84", "messageHeadline": "[WebNN EP] Use NCHW as preferred layout for DML backend (#16037)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 64, "deletions": 25}}, {"node": {"oid": "04a8f50674a4db259ffb94224068624377886b3b", "messageHeadline": "New configuration to limit the arena extension (#15983)", "author": {"name": "Yuhong Guo", "email": "guoyuhong1985@outlook.com", "user": {"login": "guoyuhong"}}, "additions": 152, "deletions": 36}}, {"node": {"oid": "60bb07307bccaa9145d1a9ba97575a85c5c38b9d", "messageHeadline": "Fix the TRT GPU build job in python packaging pipeline (#16073)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 26, "deletions": 17}}, {"node": {"oid": "cc0c5e5612024d3429b0e4bcca75aef824bdc98d", "messageHeadline": "Fix an error in test/shared_lib/test_inference.cc (#16090)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "76fd9aa745db596d5a1a8053fa76a1f0c92cd1b7", "messageHeadline": "[Fix] Some pipelines have to be using VS2019 (#16034)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 78, "deletions": 31}}, {"node": {"oid": "34fe8fb06996742d4992f8802b2061b0c1edf2e1", "messageHeadline": "Type hint for ORTModule (#15938)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 350, "deletions": 160}}, {"node": {"oid": "70d2dc8209852541cd8ec0d61e88911c9683f31c", "messageHeadline": "[DML EP] Fix issue with --dml_path build option (#15972)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "63c9973b7a8194323b00d7d2e25ba180bb9cb3be", "messageHeadline": "Fix cuda provider crash on it (#16056)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "105f5f0f20d5308ca081e47825d18d4d8737eaea", "messageHeadline": "Avoid trt deprecated api warnings shown as errors during ORT-TRT buil\u2026", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "84f1af7ff5509fd1bd5bc6e7ae2e3cb7f00bf14d", "messageHeadline": "ort build flag fix (#16072)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "20857c4ff2e36e53f4713cd418b84e88ac9337ef", "messageHeadline": "workaround test failure in ci (#16070)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "f316bc57c473f657bb9365a87bdd4c88f8735e3d", "messageHeadline": "[CoreML EP] Implement Unary & Reduce operators (#15532)", "author": {"name": "Shukant Pal", "email": "SukantK2002@outlook.com", "user": {"login": "ShukantPal"}}, "additions": 239, "deletions": 2}}, {"node": {"oid": "954ea6604aaf468db661dbfe2b6b3fbd5601e5cc", "messageHeadline": "[DML EP] Register pad18 (#15985)", "author": {"name": "Linnea May", "email": "linnealovespie@gmail.com", "user": {"login": "linnealovespie"}}, "additions": 73, "deletions": 32}}, {"node": {"oid": "bcd8b73343d0ac4e14a725cf7b94533ff64e79c7", "messageHeadline": "[WebNN EP] Upgrade max supported opset to 19 (#16036)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "efc84a43e810b05a440d7baff1b40b14e2118085", "messageHeadline": "[QNN EP] Add session option to disable fallback to default CPU EP (#1\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 214, "deletions": 16}}, {"node": {"oid": "b9d39e34053b6d595306e49ceb6f3b8bf3f4f050", "messageHeadline": "Fix cuda Transpose bug 16039 (#16042)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 42, "deletions": 2}}, {"node": {"oid": "96ee72d7f8d5e627b62c26c579dca2008aa4887f", "messageHeadline": "[QNN EP] Support Resize with 'asymmetric' transformation mode on HTP \u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 112, "deletions": 71}}, {"node": {"oid": "55c3f4b28f2a5e4118df7ef44cc8e7ebe7ae7cf4", "messageHeadline": "Fix CoreML Flatten handling of axis attribute (#16046)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 35, "deletions": 1}}, {"node": {"oid": "1b498e414f4eb106132827cbf9279120ccfdecc3", "messageHeadline": "[QNN EP] Redirect Qnn log to Ort log (#16019)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 49, "deletions": 61}}, {"node": {"oid": "b457cfaa8f2474271fe9bbbd02a0529ac1d9cfde", "messageHeadline": "Enable conditional optimization automatically (#15885)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 566, "deletions": 165}}, {"node": {"oid": "de0a973b6e328ae3271f91ce7a2dd7dfe5dcf4ed", "messageHeadline": "[Bug Fix] Incorrect comparison for FromBuffer in TrainingSession.cs (\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 11, "deletions": 10}}, {"node": {"oid": "2fddc65c8c08eb79707c172d1153380e0a111916", "messageHeadline": "[ROCm] add hipblaslt into GemmFastGelu TunableOp (#15945)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 139, "deletions": 7}}, {"node": {"oid": "684e900e960cc5fbb2fcd1dd9e80ccb43a9fc310", "messageHeadline": "Remove NETSTANDARD1.1 moniker and NETSTD1.1 specific code (#16018)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 28}}, {"node": {"oid": "d35361bf9d534ebc137ab1b817488638189de529", "messageHeadline": "Fix python pipeline for AzureEP without using root (#16023)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 18, "deletions": 24}}, {"node": {"oid": "22a578c06c34510aa41e27c5131c9ef50a7c2caa", "messageHeadline": "Use node name to uniquify the subgraph nodes. (#15855)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "4dc4470cc7e178d9a8a134e6617e09394ce41d5a", "messageHeadline": "Fix fusion for two LayerNorm sharing same input but with different we\u2026", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 88, "deletions": 1}}, {"node": {"oid": "5607a7151aa177da37d83f54339e7ef1c06cb19f", "messageHeadline": "Introduce register-efficient warp-wise Softmax (#15266)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 118, "deletions": 56}}, {"node": {"oid": "0204594f90f1ba6f9223fe29331193b24507ab06", "messageHeadline": "Cleanup WASM cmake code (#15996)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 47, "deletions": 53}}, {"node": {"oid": "e9e6bedf37428e394bf24c2e706e734748e7374c", "messageHeadline": "[js/webgpu] generate operator table for webgpu (#15954)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 161, "deletions": 3}}, {"node": {"oid": "18f17c555d51caee83f15983c2620f463fbaddd1", "messageHeadline": "[js/webgpu] fix buffer size when download (#15990)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "85cacf315ba77eb724649136951ec31a08ffc779", "messageHeadline": "[DML EP] Add MultiHeadAttention and fix Attention (#15727)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 1116, "deletions": 477}}, {"node": {"oid": "dc06c255b4842d8e291de5b5d74c9aa062b15ef4", "messageHeadline": "fix transpose optimizer on GPU EP (#15988)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 31, "deletions": 13}}, {"node": {"oid": "4324d2173b44eda7967ee2348af7c9134f8c1350", "messageHeadline": "[QNN EP] Enable Qnn context cache to save model initialization time (\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 968, "deletions": 184}}, {"node": {"oid": "4dfb89b3ad4505acc80aeda4a179ab61907e047c", "messageHeadline": "Implement mutex-free spin lock for task queue (#14834)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 65, "deletions": 2}}, {"node": {"oid": "0b0a359520504e3c51779c2891b4670fbd4894b7", "messageHeadline": "[CAPI] CAPI impl refactor (#15974)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 83, "deletions": 172}}, {"node": {"oid": "310b22aa0c1c7c7675dbd5efec4469d84bfca494", "messageHeadline": "[DML EP] Update DirectML version to 1.12.0 (#16011)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 13, "deletions": 18}}, {"node": {"oid": "d78bbf5ef20958c18db3313df0446639fd775f1b", "messageHeadline": "[ROCm] remove ROCm5.2.3, ROCm5.3, ROCm5.4 from pipeline (#16004)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 41}}, {"node": {"oid": "a74fdeb7fc480fb73952466e777ddf4541fd8547", "messageHeadline": "fix unused var warning in contrib_ops/cuda/bert/attention.cc (#16010)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "0f8e66d905e796f93da372396331bb15c5108a04", "messageHeadline": "optimization for whisper model with decoder masked multihead attentio\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 246, "deletions": 72}}, {"node": {"oid": "be6c0bb53c8cd27e7fd1041ad5cbd03c25e61030", "messageHeadline": "Update cgmanifests/generated/cgmanifest.json to fix a syntax error (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 11}}, {"node": {"oid": "0fed00c04d065d9b10c97261ae066280b40f7914", "messageHeadline": "fix topo sort in quantization tool (#16003)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 133, "deletions": 79}}, {"node": {"oid": "ea7b2deffdc67edce7c31754be206e49bf465000", "messageHeadline": "Removing C4090 warning  suppression (#15994)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 0, "deletions": 9}}, {"node": {"oid": "0c815a95b73535b2fa65598cb0b9b44340ce24ed", "messageHeadline": "android package fix (#15999)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "842b1a3472ce715ff4ce2bc62cb8b9adc389c2c1", "messageHeadline": "Revert a change in #15797: restore the correct version of emsdk (#15995)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "648bedf91a9e390fadbbabdbbb9d5239f5a9d201", "messageHeadline": "[CoreML EP] Minor changes to allow CoreML EP to handle more nodes and\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 5, "deletions": 6}}, {"node": {"oid": "5a8b892bdca6cfec3b8d05dc9da362569118c73a", "messageHeadline": "[C#] Address the concern of append EP throw (#15973)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 20, "deletions": 5}}, {"node": {"oid": "6d46007028268e135588925e83e4944cdf8eaff6", "messageHeadline": "Add explicit 'set +x' before printing a vso[] command to avoid output\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 33, "deletions": 8}}, {"node": {"oid": "d98763473a4640e574a8aaad6c6ca0dae6c4b58c", "messageHeadline": "Change CUDA pipelines to download CUDA SDK in every build job (#15915)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 31, "deletions": 13}}, {"node": {"oid": "856afa49ddbbbd0aa20b0ba6e37a834dcc9a3ea6", "messageHeadline": "[C#] Add missing rocm csharp api (#15540)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 488, "deletions": 20}}, {"node": {"oid": "0d6416c0e9eb9ba537890d0fe3f695693a08a8fe", "messageHeadline": "DML EP Bitwise operators opset 18 (#15892)", "author": {"name": "Linnea May", "email": "linnealovespie@gmail.com", "user": {"login": "linnealovespie"}}, "additions": 308, "deletions": 21}}, {"node": {"oid": "5abaca9d6969f65998708abf6cbc7616036b1769", "messageHeadline": "add maybe unused attribute to vars only used for logging (#15970)", "author": {"name": "Vrajang Parikh", "email": "vrajang@outlook.com", "user": {"login": "vraspar"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "6d43d51eb045a89ed8fe0322578b16d7015368c5", "messageHeadline": "[Fix] No test result report while not using ctest (#15976)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "2881d849d401cb9e8497c28d32e8ad78bceda425", "messageHeadline": "Update Win-CPU-2021 to onnxruntime-Win-CPU-2022 (#15967)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 26, "deletions": 31}}, {"node": {"oid": "084d0d0d2d1ee208ada791dd7be7f3b2adb702f5", "messageHeadline": "Update github issue template for 'web': add EP (#15955)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "0ff915eba83c5f520491c4b050993e758c035da6", "messageHeadline": "[DML EP] Add frequent upload heap flushing (#15960)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 18, "deletions": 0}}, {"node": {"oid": "270c09a37f36e13ee986144b2385199d23e030a0", "messageHeadline": "Add timestamp logits processor for whisper (#15853)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 254, "deletions": 7}}, {"node": {"oid": "f62f722c70cf690e40a058ab5ebaf4648ea956ac", "messageHeadline": "integrate triton into ort (#15862)", "author": {"name": "kailums", "email": "109063327+kailums@users.noreply.github.com", "user": {"login": "kailums"}}, "additions": 796, "deletions": 2}}, {"node": {"oid": "a7ad859e3ab60bddfcf2fefa96bfcb550f0fc04c", "messageHeadline": "DML EP Register Split18 (#15931)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 42, "deletions": 3}}, {"node": {"oid": "04ea561fc88433015829718923ed2e8e5ea703e2", "messageHeadline": "[js/webgpu] throw error when WebGPU=ON and SIMD=OFF (#15924)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "780442b9f6a9fdaa05452dfdd3c3c53bcb7d2bdc", "messageHeadline": "Change windows machine pools to use VS2022\u2028 (#15806)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 439, "deletions": 59}}, {"node": {"oid": "7faad536326c8fea3dfc9bd69262f5fbd7c5c470", "messageHeadline": "Set default option for package name and build arg options (#15958)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "1079df6aaa51c1d6b5386b348cfe568313fcd677", "messageHeadline": "Update StableDiffusion path after cloning repo (#15948)", "author": {"name": "Akash", "email": "akash@akashpaul.com", "user": {"login": "iakashpaul"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6b7181d31d27ed8649cece5a9298c547a3521259", "messageHeadline": "Add C# API documentation for training (and some other changes) (#15935)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 250, "deletions": 107}}, {"node": {"oid": "a0ccb95f3cabcd50ff1c85948819be083e25eb25", "messageHeadline": "add option to load pretrained weights for T5 model (#15951)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 15, "deletions": 1}}, {"node": {"oid": "e96f10d27b5fc6426c019ffd57762ea7a26d886b", "messageHeadline": "[ROCm] reduce batch size to fix CI error (#15714)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bc5018a4e16e71044e3fce531bcd8dee7ec421db", "messageHeadline": "[QNN EP] test coverage for MaxPool (#15904)", "author": {"name": "Aung T Naing", "email": "aung.thetnaing@gmail.com", "user": {"login": "aungthetnaing"}}, "additions": 319, "deletions": 0}}, {"node": {"oid": "22a9a1a630c6f171df3e6cba017737cefae13f3f", "messageHeadline": "[js/webgpu] only register webgpu backend when it's available (#15922)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dc383ed4cea3828828204304b334704aa64abdb0", "messageHeadline": "Basic CSharp packaging support for ROCm EP (#15535)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 124, "deletions": 18}}, {"node": {"oid": "204111a79ed393cf3fd9d5602376a08a22ba1a5b", "messageHeadline": "[js/webgpu] support proxy for webgpu (#15851)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 84, "deletions": 66}}, {"node": {"oid": "f3b8130d1a79b54fc6032e3e1800b873847f0b15", "messageHeadline": "[js/web] support npm run pull:wasm [buildID] (#15877)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 39, "deletions": 12}}, {"node": {"oid": "00c1da5e0a019c644f83481e849d29c363635d2f", "messageHeadline": "Fixing NhwcFusedConv fp16 (#15950)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 24, "deletions": 10}}, {"node": {"oid": "5b663d6797f194e810948e30ca03b3c20c6dc513", "messageHeadline": "Whisper Multitask and Multilingual (#15936)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 966, "deletions": 146}}, {"node": {"oid": "3418ca28a868e476000b6abad0d22b6b73ebbcf7", "messageHeadline": "pack qkv in t5 decoder (#15801)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 130, "deletions": 45}}, {"node": {"oid": "a8d9b29cd28999a2d0ec69c1a75e98f4b8828668", "messageHeadline": "support AveragePool19 and Pad19 (#15597)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 578, "deletions": 21}}, {"node": {"oid": "22bc020dcdd8e552aba60c0b4488e08816a7b178", "messageHeadline": "Ryanunderhill/ortapibase enforce (#15940)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "825d691617abbb26d30a246338d543cfef3bf919", "messageHeadline": "Unify cuda & trt version on few CIs (#15943)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "fa16e2e0f39c02afcf4f0d95ec141aa095dfa0bf", "messageHeadline": "Register CPU OptionalGetElement, OptionalHasElement on DirectML (#15926)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 80, "deletions": 13}}, {"node": {"oid": "18133ddadb5889826ecefd1f9971149d5b4b5eca", "messageHeadline": "[doc] add LeakyRelu to coreml supported ops (#15944)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "4b1d9d796ad3846468aeec5776e5a4e980485eca", "messageHeadline": "Scope down UT (#15939)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "5542e70dd161a436ad82b4fadf88d67671d11c83", "messageHeadline": "[QNN EP] Update default QNN SDK version to 2.10 for QNN NuGet pipelin\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "3b8f3a086f3be4456d2c43cc8b540d1b4ad4b179", "messageHeadline": "change the EP device to default OrtDevice() for memoryType equals CPU\u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 9, "deletions": 12}}, {"node": {"oid": "af6cb2af8726714196bab78393dbc3839ad6e873", "messageHeadline": "[ROCm] update ROCm/MIGraphX CI to ROCm5.5 (#15905)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 46, "deletions": 46}}, {"node": {"oid": "b20d5e85d5ca2333e9c0f52b595254c41c40f7ab", "messageHeadline": "Update Cuda to 11.8 in 2 Linux GPU workflows. (#15925)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 17, "deletions": 6}}, {"node": {"oid": "310273cbe4e4fd237ca1d444d4faa094bc11de87", "messageHeadline": "BeamScorer to use contiguous arrays for BeamHypotheses (#15923)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 71, "deletions": 118}}, {"node": {"oid": "896a963492fc406866a796eeb9c4a422f2b2f6e2", "messageHeadline": "Adust GetVersionString() GetBuildInfoString() signatures and move the\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 62, "deletions": 43}}, {"node": {"oid": "9fe6d588570adcab70c824f231b32de7211c0c31", "messageHeadline": "Separate execution plan serialization for a new PR. (#15916)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 0, "deletions": 50}}, {"node": {"oid": "984dd02df32927e457b4b0d117ac9d86783edd60", "messageHeadline": "Update optimize_pipeline.py to use __name__ detection (#15866)", "author": {"name": "Chester Liu", "email": "skyline75489@outlook.com", "user": {"login": "skyline75489"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "9328a0f9558f76641aa962c2e6b0b0b3156c970d", "messageHeadline": "[js/webgpu] run test on chrome instead of chrome canary for webgpu (#\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 25, "deletions": 26}}, {"node": {"oid": "143551092fea84eac84a25aa7e5740a292df90b0", "messageHeadline": "fix: setting builder optimization level to TRT 8.6 default (#15897)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 17, "deletions": 15}}, {"node": {"oid": "b473d3eee5acf5274e7c370ec53b2aeead8103c4", "messageHeadline": "Reenable ConstantOfShape TypeTests (#15910)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 3, "deletions": 14}}, {"node": {"oid": "e5189330d591d48d34cb08aa6f144141419217f2", "messageHeadline": "Address OOM Issue when exporting Whisper (#15880)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 13, "deletions": 51}}, {"node": {"oid": "000a60008026e71743fcb157b8e5b3f6626afe04", "messageHeadline": "DML EP Mark ImageScalar Test As 'won't fix' (#15894)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1bebc88069ad51fc8540967afa39724c3f73c938", "messageHeadline": "[SNPE EP] Add option to enable SNPE init caching feature (#15917)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 40, "deletions": 9}}, {"node": {"oid": "7c4e8267e74ac859729d335620bc4272e10b446e", "messageHeadline": "Implement openAI endpoint invoker for nuget (#15797)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 259, "deletions": 25}}, {"node": {"oid": "0e7ae13e7411feb6500ffd96f45987be02dfbc8c", "messageHeadline": "Run Linux GPU tests in docker container (#15872)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 34, "deletions": 26}}, {"node": {"oid": "902c5f53aebf272337d377740ba5f0c6abe14f72", "messageHeadline": "add cutlass fmha support in PackedAttention (#15838)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 302, "deletions": 67}}, {"node": {"oid": "27b2815d42430d704e979b7aea54084579d0ad2e", "messageHeadline": "Update publish-csharp-apidocs.yml .NET 5->6 (#15854)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 22, "deletions": 14}}, {"node": {"oid": "1a73d618297c080cafff0a67b85c7829a8e629ad", "messageHeadline": "Update eigen to 3.4 and remove the eigen from git submodule (#15875)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 5, "deletions": 8}}, {"node": {"oid": "7c58d013aa3715f5421f03f90690982625fdae6e", "messageHeadline": "Remove Ubuntu 18.04 usages (#15781)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 253}}, {"node": {"oid": "cf062dbdb12e11a80eeeb6f02b2be20e0f75e75c", "messageHeadline": "FlatBuffers fails to compile with gcc13. (#15787)", "author": {"name": "sdegrande", "email": "samuel.degrande@univ-lille.fr", "user": {"login": "sdegrande"}}, "additions": 19, "deletions": 0}}, {"node": {"oid": "756cf3a76f471224c088f6eac3dee99950410342", "messageHeadline": "increase web CI timeout (#15876)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e15ab78052ca9f3a63174f5fd1862d7027d4be07", "messageHeadline": "Ryanunderhill/beamsearch simplify (#15883)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 103, "deletions": 192}}, {"node": {"oid": "657ab2f43c776f93c65f9b00282a754d5e7757ca", "messageHeadline": "Sync between parent node and subgraph  (#15757)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 104, "deletions": 12}}, {"node": {"oid": "fed52053a738118da37c97a3793cd16cafd28f4d", "messageHeadline": "Refine a bit (on device training) (#15803)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 63, "deletions": 120}}, {"node": {"oid": "346ec1237733f804bfcd6f1e75935ef6694ea163", "messageHeadline": "Fix no contrib tests in TVM CI (#15895)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "19f2cc6fb6d57324057006771120d005cbce24e1", "messageHeadline": "Check the AMX tile configuration is unchanged (#15387)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 11, "deletions": 6}}, {"node": {"oid": "ac9ae9f7c599bc84747449fc0a40182bd266fe29", "messageHeadline": "update onnx release 1.14 for docker files (#15680)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 20, "deletions": 27}}, {"node": {"oid": "bd58109678fd6b19741ad91809c8d0192b9bfd4a", "messageHeadline": "relax threshold of InternalNumericalCheck on cpu from 0.0001 to 0.000\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "36c9ae0f58cec8e6f09eaab88041c693592e2d0b", "messageHeadline": "Fix  release version suffix for RC builds (#15865)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 27, "deletions": 30}}, {"node": {"oid": "95a4607dcf330c128f6b6a2ffbcedb3193f6f81c", "messageHeadline": "User/linneamay/roi align 16 (#15812)", "author": {"name": "Linnea May", "email": "linnealovespie@gmail.com", "user": {"login": "linnealovespie"}}, "additions": 451, "deletions": 12}}, {"node": {"oid": "c7b27f44866f6139ecea32b5965e90a34fca3b9e", "messageHeadline": "[DML EP] Add Outer Product Einsum (#15850)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 30, "deletions": 4}}, {"node": {"oid": "e0c1fa35a87657e3e7945fbc7422d69226531fe8", "messageHeadline": "update stable diffusion script and doc (#15846)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 92, "deletions": 56}}, {"node": {"oid": "b473e3f3c6f38f14a2bfefea55d175355a28b70a", "messageHeadline": "[DML EP] Update DirectML version to 1.11.0 (#15858)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 5, "deletions": 9}}, {"node": {"oid": "c2e631d356b8186721f506911ba08fcabf396971", "messageHeadline": "Update Conv-Add-Relu Fusion Transformation (#15834)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 110, "deletions": 18}}, {"node": {"oid": "02d94bcc8ed047ec4b1fa32c5f11f7abb728b7df", "messageHeadline": "[js/web] fix terser reserved symbols for worker (#15864)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 38, "deletions": 3}}, {"node": {"oid": "357e6289bec25491a88c66847b9780f0bafb85b3", "messageHeadline": "[wasm] allow pull debug artifacts from script (#15859)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 18, "deletions": 14}}, {"node": {"oid": "475f661acdbabffdac4fc97d01922a623d5f7611", "messageHeadline": "use __hmul2 instead of __hmul2_rn (#15852)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "34cb293c6b31d91ea94dd8fed4fddc36cf366b7c", "messageHeadline": "Remove unused ADO YML pipeline template (#15857)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 0, "deletions": 789}}, {"node": {"oid": "686d42e6c832f3ddd6fc27a1ba26c1cfd6555fbd", "messageHeadline": "layer_norm_fix (#15844)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1022, "deletions": 800}}, {"node": {"oid": "00b1e79e048e36e404d5b6516238d347aa5ba81f", "messageHeadline": "Support WebNN EP (#15698)", "author": {"name": "Wanming Lin", "email": "wanming.lin@intel.com", "user": {"login": "Honry"}}, "additions": 3404, "deletions": 21}}, {"node": {"oid": "685e5b00f603ed67e50dc9e876baf5d150a5dfcc", "messageHeadline": "NhwcFusedConv: Add before Activation (#15837)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 37, "deletions": 34}}, {"node": {"oid": "003c7d3e4dea49e3fd5bd28db34f043874b1f74f", "messageHeadline": "Add CPU allocation test for multiple GPU distributed run (#15829)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 99, "deletions": 0}}, {"node": {"oid": "817d70a63b62dcab40c2fd038ca0dc30072cb5be", "messageHeadline": "[js/rn] Fix extensions header include issue (#15800)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 24, "deletions": 11}}, {"node": {"oid": "0457fd0b4078fbca07d9506aa1bcda2a839afc81", "messageHeadline": "upgrade emsdk to 3.1.37 (#15817)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 95, "deletions": 19}}, {"node": {"oid": "191ee1d3c0ddcc6d7c97458a30e7d3ec9c0a079f", "messageHeadline": "Fix symbolic shape infer empty value_info (#15842)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 18, "deletions": 12}}, {"node": {"oid": "045c623415b7cc2057c876c4173d83d320fe04d2", "messageHeadline": "Make Nuget workflow easy to debug (#15808)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 94, "deletions": 26}}, {"node": {"oid": "5e9b42326c0ce5523176211834fdb3bc8142fd0c", "messageHeadline": "Fix packaging pipeline for nightly builds (#15839)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 19, "deletions": 6}}, {"node": {"oid": "2735e0d031b80b66b3d90d0580a14e2e2058da23", "messageHeadline": "[ROCm] simplify ck data type Adaptor (#15734)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 55, "deletions": 80}}, {"node": {"oid": "42d62b8f2b69f9cdfd8a650a9e330b2c547fe20c", "messageHeadline": "Fixes to get stable diffusion benchmark running (#15755)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 14, "deletions": 13}}, {"node": {"oid": "41457885e012c42556a7b8d7f1294115a2768dc2", "messageHeadline": "[ROCm] add rocm5.5 to python package pipeline (#15820)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 16, "deletions": 4}}, {"node": {"oid": "ed31e4b737283b52078bbdf9d8025b9dd7f60634", "messageHeadline": "Add nuget release version suffix to support publishing rcs to nuget.o\u2026", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 60, "deletions": 18}}, {"node": {"oid": "dfac09650193287574bf9af8013e299e6525b4db", "messageHeadline": "Fix segfault for multiple GPU run (regression) (#15823)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 39, "deletions": 38}}, {"node": {"oid": "2b7f26af7c097294c9ce6aead3399a3af3b112d0", "messageHeadline": "Add GridSample implementation to DirectML (#15788)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 218020, "deletions": 19}}, {"node": {"oid": "45f5c276320e46ee26304cbfe9311327f69bc598", "messageHeadline": "[QNN EP] Update default QNN SDK to version 2.10.0 (#15818)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 189, "deletions": 44}}, {"node": {"oid": "2f6df99b89afc925647f7751f54e6b0f00de02fc", "messageHeadline": "Register fp16 FusedConv to CPU EP (#15824)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "5a43828b3d73028bfd33b3856f82698d9ab02cb1", "messageHeadline": "update ort extensions to 94142d8391c9791ec71c38336436319a2d4ac7a0 (#1\u2026", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 13, "deletions": 53}}, {"node": {"oid": "41a19ae1b5d381b21b5ea6938cc7b77fb33be16d", "messageHeadline": "[js/webgpu] fix Transpose with non-float tensor (#15819)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e139ae238bbce6772be76fcb7ea13fb402adc483", "messageHeadline": "Add a codesign step to Windows AI nuget pipeline (#15816)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 25, "deletions": 0}}, {"node": {"oid": "d1b2b35cd2fd451b5c09afaef74f056773286e5b", "messageHeadline": "Various fixes to the CSharp setup (#15782)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 147, "deletions": 39}}, {"node": {"oid": "272aab4afa92fedca5b537c78667d344a57b2325", "messageHeadline": "Fix issues on Windows for Vitis AI (#15810)", "author": {"name": "BoarQing", "email": "yuz75@Pitt.edu", "user": {"login": "BoarQing"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "412d05a1d256cd63dddefa409d3213275fe62fc2", "messageHeadline": "[ROCm] Update cmake  (#15807)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "58ee076750ee05b33f787f87648df8ae25591ec4", "messageHeadline": "Bump engine.io from 6.4.1 to 6.4.2 in /js/web (#15799)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "4712009f8a83592d6da5ea513ac482d6b666cfb7", "messageHeadline": "[js/web] add target ort.webgpu.min.js (#15780)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 84, "deletions": 21}}, {"node": {"oid": "8e610f25d8a14d6a4f311106d15d90d4f85aef91", "messageHeadline": "Implement lite custom op API  (#15778)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1072, "deletions": 106}}, {"node": {"oid": "34fcdd83c80976ee14f67bcbdc5515520d89c07c", "messageHeadline": "Update softmax_grad_impl.cu: add constexpr (#15794)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "df7424e11fb55ac81efcbc8850f0ac89b8082a81", "messageHeadline": "[JSEP] fix constructor for OrtDevice (#15805)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "33d1372729e329027823d1dedc036c76d8a0683d", "messageHeadline": "[wasm] revert emsdk to v3.1.19 (#15793)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 6}}, {"node": {"oid": "422606ea76c05d2b1d07e775a49faf93a36f055c", "messageHeadline": "Bump engine.io from 6.4.0 to 6.4.2 in /onnxruntime/test/wasm (#15798)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "e464588a0e1be83cff7d61bfddca4a2d62a48520", "messageHeadline": "Avoid generating training documentation during packaging (#15795)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 0, "deletions": 42}}, {"node": {"oid": "5eedd884c84404478d4bb411678c4b238bd4a277", "messageHeadline": "Adding support for conv fp16 fusion on Resnet50v1 (#15474)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 250, "deletions": 2}}, {"node": {"oid": "1fb2f2605b83fe29f047934cea9b33bd9f3c3e76", "messageHeadline": "Update VERSION_NUMBER (#15773)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 31, "deletions": 101}}, {"node": {"oid": "ba7b83ff3c48fa5df931001ffe1b96daf14d1c5d", "messageHeadline": "Remove onnxruntime_PYBIND_EXPORT_OPSCHEMA definition from onnxruntime\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9, "deletions": 31}}, {"node": {"oid": "41c082fdded91dcbf4f0756cc63e1363746c70e7", "messageHeadline": "Add a Github workflow for Prefast (#15763)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 63, "deletions": 6}}, {"node": {"oid": "d53324d4a72bf76518bede6562656826c1da8610", "messageHeadline": "Update cmake version in a few places (#15775)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "e3ec2b3a8e4ab818203f3f5a423d314186244bd6", "messageHeadline": "Exclude cases from reduced build (#15779)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ef1f17f3dcf15995283a0358fc9c9f54d337a256", "messageHeadline": "[wasm/JSEP] add threaded build to artifacts (#15777)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "2d519d21af3389a195c89e8b68d00377879fc714", "messageHeadline": "Python documentation for onnxruntime-training (#15765)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 329, "deletions": 150}}, {"node": {"oid": "abdd4f518a144035fee3b369996d8416a024bdaa", "messageHeadline": "Update TRT Windows Cuda 11.6 to 11.8 (#15746)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "328cabb194e6f4094ee677b5199896fc167e2581", "messageHeadline": "Download protoc from Github Release instead of Nuget (#15731)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 53, "deletions": 77}}, {"node": {"oid": "e901cdbf544228278a08fc545d1fbf99bf40273c", "messageHeadline": "Add wildcard paths to the API docs generation workflows (#15313)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "4b8025e4929ec92c8751e472e3bb8f1a878dcc08", "messageHeadline": "Parallelize fp16 pooling operators (#15766)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 17, "deletions": 13}}, {"node": {"oid": "bc58fd541318f6f5197c7f5750c3a12291faa04f", "messageHeadline": "fix compilation error in no absl build (#15769)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "92309187b3e736ddb5031a7e63eed42a89ed6f2c", "messageHeadline": "Allow compilation using clang when using cuda. (#15672)", "author": {"name": "Sohaib Iftikhar", "email": "sohaib1692@gmail.com", "user": {"login": "sohaibiftikhar"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "034698cf6a3f9a7365025ddbe6a8b5fc17faa682", "messageHeadline": "Revert \"Implement lite custom op API (#15590)\" (#15768)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 104, "deletions": 1069}}, {"node": {"oid": "090312af711ddae1a3936c8dde180551b3ccefbb", "messageHeadline": "add local state dict option (#15759)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 18, "deletions": 2}}, {"node": {"oid": "391f897983452aa28ebb39d879c3eac7f1262986", "messageHeadline": "Bring back SLN cuda kernel and use provider options to switch to stan\u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 398, "deletions": 32}}, {"node": {"oid": "921961547118901a07ccfd761f93cd63866221fb", "messageHeadline": "Fix python AP docs generation (#15760)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 2, "deletions": 6}}, {"node": {"oid": "5352f6d9b05f1c1174e2e9c7a9b726adf32e25cc", "messageHeadline": "Make \"--cuda_version\" build arg optional (#15758)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 19, "deletions": 10}}, {"node": {"oid": "7f293d065a54099b373f6c785cdfcff2341ab710", "messageHeadline": "Remove some D2D memcpy in T5 BeamSearch (#15706)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 248, "deletions": 31}}, {"node": {"oid": "0ffae8073b02f8c3e1392f506667ad79498ab93e", "messageHeadline": "Creating Nuget and Android packages for Training (#15712)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 686, "deletions": 61}}, {"node": {"oid": "4c4f688a9347c6ef38b0008725f528f92f04842a", "messageHeadline": "[DML EP] Fix dml_external_project (#15656)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "62fc6ed5a88862e9ba44ff01b05711dd60bb2aa5", "messageHeadline": "[Feature Request] Support Resize opset 19 (#15633)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 113, "deletions": 21}}, {"node": {"oid": "d58fa9805b7da86191298b4a632790a1b7a9365c", "messageHeadline": "ExecutionProvider API refactor - replace OrtMemoryInfo with OrtDevice\u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 255, "deletions": 192}}, {"node": {"oid": "bb33285ec294da2e702bd8c9285a7bfeca0114bf", "messageHeadline": "C# training api updates for on device training (#15720)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 556, "deletions": 52}}, {"node": {"oid": "c10a6a9d178446b9d1cffc42fee985bf4404ff1d", "messageHeadline": "Tensor <--> image - Adding per channel compute for Norm mean & Bias (\u2026", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 247, "deletions": 87}}, {"node": {"oid": "0770cf36999b0f20a10b39d666af241fc6d93207", "messageHeadline": "Remove C# SessionOptions.RegisterCustomOpsUsingFunction. (#15754)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 0, "deletions": 87}}, {"node": {"oid": "cdf4fc49fc0a8dfe3f2962c78ebf4b513911ca68", "messageHeadline": "Implement lite custom op API (#15590)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1069, "deletions": 104}}, {"node": {"oid": "0e9472d39159846a3e107b4861e89c19fcb9e61d", "messageHeadline": "NHWC graph optimizer (#15724)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 461, "deletions": 43}}, {"node": {"oid": "d35850c142925ad35214bb3cfe82de408fca6bf6", "messageHeadline": "[VitisAI]Update VitisAI EP to be compatible with VitisAI 3.5 (#15673)", "author": {"name": "Chunye Wang@AMD", "email": "Chunye.wang@amd.com", "user": {"login": "wcy123"}}, "additions": 2494, "deletions": 640}}, {"node": {"oid": "3df3a851143b0eb35d16acb01d75991c677796d3", "messageHeadline": "Default kOrtSessionOptionsDisableQuantQDQ to 1 when the DML EP is reg\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 23, "deletions": 3}}, {"node": {"oid": "10dff4f6657dff450f31a2f904291a959552aadf", "messageHeadline": "only add type info from symbolic shape inference for fp16 conversion \u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 47, "deletions": 7}}, {"node": {"oid": "6e652d05541ee49d2cb6acca49a529ace8c3ad35", "messageHeadline": "Support explicit TRT profiles from provider options (#15546)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1221, "deletions": 279}}, {"node": {"oid": "31e7d3d7d4532aab9ac27055396f2eca1c0fed51", "messageHeadline": "Disable TestRegisterCustomOpsWithFunction on Linux (#15747)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 21, "deletions": 17}}, {"node": {"oid": "176161348efd87104b7e95f58ea69c979027a265", "messageHeadline": "Revert \"make nuget workflow easy to debug. (#15693)\" (#15744)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 28, "deletions": 64}}, {"node": {"oid": "7ae01cec157f6687b63346c06298a75d06a4bbf4", "messageHeadline": "Update wheel path to Whisper custom export script (#15739)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "4fbc08e3c28ee20bfbb06aceb0e00d6f635535c3", "messageHeadline": "VPUX config fix and dynamic_shape bug fixed. (#15737)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "d32c540b2d1f3c0b60f36da1bc8f59f4c99cc29e", "messageHeadline": "[QNN EP] Support LRN operator (#15741)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 352, "deletions": 5}}, {"node": {"oid": "65020d433eb1a18a65ad6c5b9c080a9c872b54a2", "messageHeadline": "Prefast fixes for CUDA EP (#15726)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 116, "deletions": 100}}, {"node": {"oid": "ec2f038c6df4d71277f8d2d9a0a8954174258c96", "messageHeadline": "Update Nuget pipeline's Linux CUDA job to cuda 11.8 (#15516)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c8bd34f9750ebc0cbd2443a90ad06a647ea5c203", "messageHeadline": "[js/rn] Package dependency change to manage ort-extensions for react_\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 109, "deletions": 2}}, {"node": {"oid": "41dcf0d32e0e4cc9da7a3840c56c0a096c277d94", "messageHeadline": "Expose build information in dynamic lib (#15643)", "author": {"name": "Yuhong Guo", "email": "guoyuhong1985@outlook.com", "user": {"login": "guoyuhong"}}, "additions": 73, "deletions": 22}}, {"node": {"oid": "191deb4235b4053845c929907910a48dd32751e2", "messageHeadline": "[QNN EP] Nuget package (#15711)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 268, "deletions": 5}}, {"node": {"oid": "6a6091a5192936f67d1e7508575a9274808fb0fe", "messageHeadline": "[rn] Add support for loading model from buffer on iOS  (#13802)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 101, "deletions": 31}}, {"node": {"oid": "fe1ddd7b61c5c715f9a4678b994c844819d90c47", "messageHeadline": "Fix bug when adding Whisper to wheel (#15708)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 53, "deletions": 0}}, {"node": {"oid": "2802c547a188192b47fe76e82eef21c5c327de67", "messageHeadline": "update OnnxMl.cs (#15702)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 279, "deletions": 151}}, {"node": {"oid": "29d13cea4216497525614cc1803fe5b164ab32d4", "messageHeadline": "Cumulative update on optimizers and tests (on-device training) (#15499)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 554, "deletions": 262}}, {"node": {"oid": "8a1a40ac63f7370341ef907a8bfa78d8b3c5fe5c", "messageHeadline": "[Java] CheckpointState AddProperty & GetProperty support  (#15730)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 510, "deletions": 8}}, {"node": {"oid": "be08b47e7bc08285b48a119f1b756de67e593ce9", "messageHeadline": "Refine cast optimizer for safety (#15658)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 242, "deletions": 78}}, {"node": {"oid": "c415bc725f5e30e094d99d3d2d2cbcf9b4cfda93", "messageHeadline": "Add 'name' key to xcodebuild 'destination' option. (#15690)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c401cf4b5163bc5dd0d60970de40d4067febcf75", "messageHeadline": "Fix issue there 9573-quantizing-distilbert-models-after-optimizing-wi\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 74, "deletions": 11}}, {"node": {"oid": "7e6331d5c7ce6e9b57516afb9e333097d449f941", "messageHeadline": "Add ability to register custom ops from ORT extensions nuget package \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 266, "deletions": 70}}, {"node": {"oid": "94c9a31f83d2b2242103d24b0c540a7a0c3e9369", "messageHeadline": "[js/webgpu] fix download failure due to buffer change (#15723)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "2c3697be00dfe074d2368637c4830cca89957c04", "messageHeadline": "User/linneamay/reduce 18 (#15701)", "author": {"name": "Linnea May", "email": "linnealovespie@gmail.com", "user": {"login": "linnealovespie"}}, "additions": 45, "deletions": 14}}, {"node": {"oid": "5b826b1bc390db60315536cddeee6d32bdcd7b90", "messageHeadline": "Update cmake version in Linux build (#15707)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 35, "deletions": 35}}, {"node": {"oid": "9db24f8fec40c047044094441c90bedfe05bae6f", "messageHeadline": "Update kernel registration validation to allow kernel registrations t\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 149, "deletions": 51}}, {"node": {"oid": "39d6d7050df58c4c44d9b97ec960c7a8f0a809f7", "messageHeadline": "Change EmbedLayerNormalization mask index output to optional (#15526)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 116, "deletions": 31}}, {"node": {"oid": "d471432e10c921c32961000349b0a1565ca30d4e", "messageHeadline": "[js/webgpu] fix attribute cache key for 2 operators (#15710)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c0116af619b7bc2eecec860c643853eba1b37cbf", "messageHeadline": "[js/webgpu] operator Exp (#15713)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "627f5c97674e5af25be62f83bf98e542ad9dcd57", "messageHeadline": "support allgather on different axis (#15610)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 94, "deletions": 22}}, {"node": {"oid": "5bde1e8e37a24a2b70b69b42e76f1c6e34e35385", "messageHeadline": "Add Bluestein Z-Chirp Algorithm to DirectML DFT implementation (#15686)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2346, "deletions": 712}}, {"node": {"oid": "be5c582e65484cbd06be2b3c3e70040f20accaa7", "messageHeadline": "[QNN EP] Update to QNN SDK 2.9.0 (#15709)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 18, "deletions": 17}}, {"node": {"oid": "9773e76c446999a7c4cc574852248bfdfcd9d9b4", "messageHeadline": "Single-schema-multi-kernel (#15184)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 581, "deletions": 106}}, {"node": {"oid": "d3d232b047f85b721a0520bebbfc46de4c2c2766", "messageHeadline": "Rename onnxruntime-Linux-CPU-2019 machine pool (#15691)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "a957a872d3df3789f672e8766801755b104e692f", "messageHeadline": "Patch fix for the newly added TRT EP provider options (#15687)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 65, "deletions": 1}}, {"node": {"oid": "d3e8d7a70d3588a68748938fce50a9584cd8c1de", "messageHeadline": "Better support for cmake 3.26 and Windows ARM64 (#15704)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 21, "deletions": 1}}, {"node": {"oid": "2e1f92a9862493073b8ab0ecd7a9ef86373734f6", "messageHeadline": "Fix EP Perf pipeline (#15507)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 27, "deletions": 9}}, {"node": {"oid": "8cda1ffa28ff4033249eb5aeaa284c5dd2a10758", "messageHeadline": "Fix error in post-merge pipeline (#15717)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a952419674aa039f89f6e909e54877285f0af2dd", "messageHeadline": "[ROCm] Fix FusedConv to stop caching fusion args (#15671)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 15, "deletions": 29}}, {"node": {"oid": "2efb75bfe9bfce635b9ab76d92beafdb8e54446d", "messageHeadline": "Fold shape related operation (#14936)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1483, "deletions": 0}}, {"node": {"oid": "53ff50d19a8a49e45fba33791acaf5eee5cbaeaa", "messageHeadline": "make nuget workflow easy to debug. (#15693)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 64, "deletions": 28}}, {"node": {"oid": "926ae7d7862addd544b5273c75a0fa0895ab54ee", "messageHeadline": "Add updated skipped test for multiheadattention Packed KV & QKV (#15587)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "e63bb5acefa2c931d66566013278ee9a1a0e2c7e", "messageHeadline": "Fix a memory leak in QGemm (#15703)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 9}}, {"node": {"oid": "740d553c42f75f4778a08665eb40bdc760a4ab51", "messageHeadline": "[rn] Reland support loading model from buffer for Android (#14514)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 817, "deletions": 788}}, {"node": {"oid": "a02c885f861b7bf18d5caeebb6308bcfec08b89b", "messageHeadline": "[js/webgpu] add implementation of Relu, LeakyRelu and ThresholdedRelu\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 57, "deletions": 12}}, {"node": {"oid": "76ddc92fbde6547cff153c1e6ef81c058a52b994", "messageHeadline": "Enable RUFF as a formatter (#15699)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 10, "deletions": 34}}, {"node": {"oid": "d7ba9814cffa9615b1c525c05f094527ef28f4d5", "messageHeadline": "[prefast:Warning]: C26409 ('PackedAttention<onnxruntime::MLFloat16>::\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "97c4cab6b7e7b1f5ad824a41238c427b443a972d", "messageHeadline": "[DML EP] Massage SkipLayerNorm axes to better target metacommands (#1\u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 45, "deletions": 19}}, {"node": {"oid": "4c7b5032dad8c44799e90a003ab20d852b196bc0", "messageHeadline": "[QNN EP]Support unpack initializer from external data source (#15694)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 52, "deletions": 41}}, {"node": {"oid": "28985c47b7092a88b73e4a91d2cb1291857f0954", "messageHeadline": "[TensorRT EP] Unleash opset16-17 onnx model tests (#15657)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 20, "deletions": 3}}, {"node": {"oid": "cfb8c0e2ca99acb693b084d409f6848c8489488f", "messageHeadline": "Add Whisper custom export to wheel (#15685)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "d701dcd0279a0ffef0e317426bfc2a0e1337f6b4", "messageHeadline": "Fix Linux MultiGPU TensorRT CI (#15697)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "0ecfe839328a2da8c37d446039db958d05428e67", "messageHeadline": "[ROCm] add beam search support (#15625)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 34, "deletions": 5}}, {"node": {"oid": "699c9a520bed465bb9825054f62594c80a14b19b", "messageHeadline": "Fix TVM pipelines (#15653)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b98317b90713ad082c3b951270f9ecdcf6f3fa0f", "messageHeadline": "[js/webgpu] following up for JSEP/WebGPU code cleanup (#15666)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 35, "deletions": 351}}, {"node": {"oid": "ebaafac3f5a1e318488a1636686bb9bb911d2db8", "messageHeadline": "Openvino ep ort 5.0 (#15626)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 704, "deletions": 1301}}, {"node": {"oid": "b1b6e5522e40a26977d0a0bd0ecb8fb48bd23e11", "messageHeadline": "Update cuda 11.6 to 11.8 for Windows pipelines (#15684)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "db6a9bc033b4e228c32224faf65a63592ae96a12", "messageHeadline": "support latest deepspeed version for optim (#15682)", "author": {"name": "Rui Ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3dc9720cfc9df7e11359ddbae77a0cb00f71f904", "messageHeadline": "[QNN EP] Enable Qnn EP op support Elu, HardSwish, Atan (#15681)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 72, "deletions": 14}}, {"node": {"oid": "1524f73a09538c8488368d15e589e922a3190078", "messageHeadline": "Implement two easier random tensor generator (RTG) for flaky tests (#\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 148, "deletions": 0}}, {"node": {"oid": "f44f6c5b2e2e4eae7e637b24d118d770f5902a09", "messageHeadline": "Fix Prefast Errors (#15651)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 19, "deletions": 18}}, {"node": {"oid": "4c3e350a6ade3709359f4d804ab130161795bb29", "messageHeadline": "fix ORTModuleONNXModelException fallback OOM (#15523)", "author": {"name": "Rui Ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 59, "deletions": 5}}, {"node": {"oid": "d30831d829e22124906cac83cde766e10d6b7ae1", "messageHeadline": "[js/webgpu] make `RunFunction` return `void` (#15669)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 53, "deletions": 83}}, {"node": {"oid": "2fa10fb8033561a2021f161a7f9bc30b15d4ad9e", "messageHeadline": "Fp16 onnx pool operators, relu, leakyrelu (#15498)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 901, "deletions": 47}}, {"node": {"oid": "9bf08bdb525119b2db9453fc1eaf4275f8fe8306", "messageHeadline": "Fix iconv link issue (#15592)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "d05777ddb673575ec555b95b19b597797f664b2f", "messageHeadline": "stabilize fusion script with a seperate create_attention_node() (#15670)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 109, "deletions": 4}}, {"node": {"oid": "5885abfb35c2288d87b7676fe02bff33593a70f4", "messageHeadline": "Training Documentation (#15612)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 511, "deletions": 156}}, {"node": {"oid": "d00197aaa77c84bbba53c0619ef30a99320c2dbd", "messageHeadline": "initialize cache_indir explicitly in beamsearch with encoder decoder \u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 38, "deletions": 7}}, {"node": {"oid": "e1755541cce8e08cbb82de7527413fdc807cb810", "messageHeadline": "Fix TRT timing cache test (#15588)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 31, "deletions": 38}}, {"node": {"oid": "d0c3f92ec6bd3f370ce78abf330620781591bf9d", "messageHeadline": "[DORT] Fix fake tensor problem cuased by PyTorch change (#15664)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "3440d3a08e974aa9f7b586316220f7ae39202256", "messageHeadline": "remove 'lib/' from .gitignore (#15613)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "124ea0a801097337daad19d488031047b3cb1d44", "messageHeadline": "remove compute optimizer from lte (learning on the edge) builds (#15637)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 33, "deletions": 19}}, {"node": {"oid": "14cc02c65ccd946b2f231756e86afa76095bb1cd", "messageHeadline": "[js/web] WebGPU backend via JSEP (#14579)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 8091, "deletions": 394}}, {"node": {"oid": "8dd32fed47bb30336170429ca6ac86230901e5b6", "messageHeadline": "[TensorRT EP] avoid excessive library load/unload overhead when runni\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 36, "deletions": 37}}, {"node": {"oid": "c2acf69d13331829e0eeb4fa3c1a5d34275906a8", "messageHeadline": "support new include,lib dir structure in upcoming QNN 2.11 (#15605)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "ccb2243ee7b2689185b0f6549690682f72c6f3b6", "messageHeadline": "Update build option for training in java to enable_training_api (#15638)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 10, "deletions": 8}}, {"node": {"oid": "686fd3c22a2b2961cbbf1730d7017eda5dd61858", "messageHeadline": "Fix cuda 12.1 windows Build (#15614)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1009, "deletions": 853}}, {"node": {"oid": "dc53ddef7a2fbd242a3e10b38a45463557060d33", "messageHeadline": "Create a new C API KernelContext_GetAllocator() for Custom Op scenari\u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 60, "deletions": 4}}, {"node": {"oid": "a8e283305038c7891db437d0646a58fed52e6583", "messageHeadline": "[QNN EP]Unblock Qnn EP for Csharp support (#15640)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "c82bebde6a2b56fe9cf97de1d69e78ef96326535", "messageHeadline": "Fix the TestCUDAProviderOptions test error (#15649)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "9df1a5e605c79b7cc33c5d0d46cee040d95f1744", "messageHeadline": "[ROCm] enable LayerNorm opset Ver17 for ROCm EP (#15601)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 8, "deletions": 13}}, {"node": {"oid": "45c82eefb41cffc976c34fb2a758c8bdccfc8de9", "messageHeadline": "[OneDNN] Fix poolgrad bug (#15557)", "author": {"name": "Erick Mu\u00f1oz", "email": "erick.munoz.alvarado@intel.com", "user": {"login": "eralmual"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d1354dcc832b23d5bf198cc418f119f42639e7ba", "messageHeadline": "[ROCm] Add stable diffusion benchmark results for MI100 (#15646)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 18, "deletions": 1}}, {"node": {"oid": "8297148bde5ac8e6b800686daff98f95679c6fa0", "messageHeadline": "[ROCm] Update benchmark for stable diffusion (#15602)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 232, "deletions": 123}}, {"node": {"oid": "9e44248bf9a73db490672c9cda890fa38b4256e7", "messageHeadline": "Workaround ROCm global pool (#15481)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 411, "deletions": 0}}, {"node": {"oid": "fd6ecc3909ab1a371cf4d8c94e16957e3fcc0f84", "messageHeadline": "Add env to the TrainingSession constructor (#15635)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 15, "deletions": 10}}, {"node": {"oid": "fab3e3310562d36842182790fade767a4ddd9c9f", "messageHeadline": "[Qnn EP]Enable Gelu op support (#15631)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 99, "deletions": 172}}, {"node": {"oid": "0080bb03317946a2ec3777d0a636581e124ac17f", "messageHeadline": "Add NCHW transpose for GroupNorm (#15634)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 46, "deletions": 10}}, {"node": {"oid": "b49d428299826c58d13de893ef96ebf7f39c2472", "messageHeadline": "[DML EP] Add missing newline to image test logging (#15596)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5a675d91137c79375e29c46499d19d7289fa6e9f", "messageHeadline": "Disable random failing DML image batch test (#15624)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "633dec0b17ae99b2817271d5b1d13ae8c6147ec3", "messageHeadline": "refactor some code (#15566)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 485, "deletions": 1006}}, {"node": {"oid": "b5a1941835bc2618d616741ee59d1a776d8a0bb3", "messageHeadline": "C, C++, Python, C# API update for on device training (#15518)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1381, "deletions": 537}}, {"node": {"oid": "a6d6e45be269ae198f83906c93cb251c5dd5fb42", "messageHeadline": "Tune block size for layer_norm considering #rows and GPU resource (#1\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "2cb3fb18b5f168550ccfb607286b4f42750a76e6", "messageHeadline": "Integrate React Native E2E test with detox framework (#15133)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 2335, "deletions": 558}}, {"node": {"oid": "f3d04cd1be6844d415ef461e9ca62b518e177bc0", "messageHeadline": "[QNN EP] Update Windows ARM64 pipeline to use Visual Studio 2022 (#15\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 145, "deletions": 73}}, {"node": {"oid": "84746a8efe218d8adf0004350e31000e71bd915d", "messageHeadline": "Revert \"Retry the step of Start Android simulator (#15584)\" (#15620)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "3de33e00c709cc4340acc438e177074d9af1087d", "messageHeadline": "Fix issues for Whisper export with beam search (#15619)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "9011613b65ea7016a8c59c485ec30e6dcb86f703", "messageHeadline": "Add Trilu and GatherND to the list of supported OPs for MIGraphX EP (\u2026", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "a2f80a006be42133a9d8d2b0325cb441a4a50f4d", "messageHeadline": "update target framework to dotnet6.0 (#15615)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "6cf080ccbfa34ace14d051f12f1a08635772f86d", "messageHeadline": "Temporarily disable two tests for TRT EP (#15578)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "dfa06bf81b9f2c5bf31d1142299f80cbcaca3900", "messageHeadline": "Add link to doc for lintrunner in CI (#15604)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a5dec8eedfeedf7942fc720b304cee85d26c2068", "messageHeadline": "[C# ] Improve string marshalling and reduce GC pressure (#15545)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 684, "deletions": 470}}, {"node": {"oid": "373f912e51342c2966189b68a58f97cee5f94449", "messageHeadline": "add quantization support for whisper (#15589)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 58, "deletions": 16}}, {"node": {"oid": "4b74cb17419a08dce2998456fda9c19e057de885", "messageHeadline": "Make docker command fail if bash command fails. (#15564)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "46210556f0e34887a7fc1dc6d8a60fe48f6cec53", "messageHeadline": "BatchnormInternal avoid setting num_channels if input shape is not kn\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "11b0a18de68893222f75a14785c1cd36b04b2e5f", "messageHeadline": "Add support for cuda 11.8 and python 3.11 for training (#15548)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 44, "deletions": 32}}, {"node": {"oid": "1f7c2f724ffa6362f4d7f3f4360531458bd70d82", "messageHeadline": "Fix lintrunner configurations (#15586)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 40, "deletions": 10}}, {"node": {"oid": "9df96c7d5bf7590c5ed4c9febb37bc51e532dedc", "messageHeadline": "[QNN EP] Fix shape inference of NHWC Resize (#15477)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 207, "deletions": 42}}, {"node": {"oid": "446c478fbd87c85a0815f14a3b2ff055000cd7d5", "messageHeadline": "Add iOS Swift Package Manager support (#15297)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 270, "deletions": 37}}, {"node": {"oid": "64b63921a27d67e6a6c371308d7925669de1f172", "messageHeadline": "Retry the step of Start Android simulator (#15584)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "5b6f79e79b68912cc0b1a7f38b2e935027e98bd0", "messageHeadline": "Improve windows build cache steps (#15537)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 59, "deletions": 37}}, {"node": {"oid": "29d00fb7767b9202fdda01cbd64f5150e5427eb4", "messageHeadline": "Set proper default values for pool attributes (#15559)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f2889b41c1a3f6b9197687949983fed6218c791e", "messageHeadline": "[AMX] Update assembler check (#15501)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "142220ad87a609f95bdf7df157eca108a887c808", "messageHeadline": "Fix cmake 3.25 debug info config (#15565)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "573e4cf95fba64452deccb21bdd8b923ca57146b", "messageHeadline": "[Fix] Python Packaging Pipeline exception. (#15568)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "59ea35d5929bcaa50977c57481a88a1f82fd9894", "messageHeadline": "[ROCm] add CK GroupNorm to GroupNormTunable (#15510)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 676, "deletions": 66}}, {"node": {"oid": "a66af390fa3688f6185e6a47861cd81afbf15375", "messageHeadline": "[C#] Allow passing various options when creating singleton Environmen\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 857, "deletions": 283}}, {"node": {"oid": "6115c8fd1f4c32e058ac7a841fdddb0e99e509cc", "messageHeadline": "Add TRT plugins support using custom ops (#13847)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 478, "deletions": 7}}, {"node": {"oid": "cb83d2b1a97ef06cb3fbbb9779a0cbf157a9fdde", "messageHeadline": "[js/web] allow script to use partial success build (#15547)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "901c2bc38452a1cbeefd5544eee55165ba6e467d", "messageHeadline": "Whisper Model Optimization (#15473)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 4509, "deletions": 326}}, {"node": {"oid": "53d304d4d26d84dc6a3c7c11050e301c37752329", "messageHeadline": "optimize gated gru cuda kernel (#15525)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 211, "deletions": 41}}, {"node": {"oid": "831734a46e8139c52ed54b40288c8de89c0813b9", "messageHeadline": "Fix lint errors missed due to new commits (#15558)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 17, "deletions": 18}}, {"node": {"oid": "698e9f71cd72ed6176938ff093b40e57f92cd681", "messageHeadline": "Improve cache hit rate in windows build (#15538)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 16, "deletions": 29}}, {"node": {"oid": "cf19c3697d126982acf45c040729269e53b5d9f4", "messageHeadline": "Run clang-format in CI (#15524)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 11304, "deletions": 11120}}, {"node": {"oid": "2700d01642a534dc3e57e0da5b304e25a55771aa", "messageHeadline": "Add Bluestein Z-Chirp CPU EP implementation for the DFT operator (#15\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 121, "deletions": 33}}, {"node": {"oid": "919d8f2660de189e7b63f1c069bb7e5e51d3b111", "messageHeadline": "update with onnx main (#14929)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 289, "deletions": 82}}, {"node": {"oid": "d8dfda2e08e06608d27cf7fa80f2caf83a2cf34e", "messageHeadline": "Minor fix for differently scoped cpu_ep usage (#15550)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "8bec6cd029b9c57f78e40ac60d5595f12de65f40", "messageHeadline": "Refactor FusedConv test (#15512)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 87, "deletions": 119}}, {"node": {"oid": "9d26f8f4fe1654bfe7e2f631f892a8cf08caf050", "messageHeadline": "Use os.fspath on Path (#15530)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 2, "deletions": 5}}, {"node": {"oid": "a30b57da6e1d985a5d6ecf433206c212cc469f8c", "messageHeadline": "Fix/Enhance convert_generation tool for SkipLayerNorm, op_block_list.\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 56, "deletions": 17}}, {"node": {"oid": "a36caba073cccc8fa3390e9e118fa9065d8fd6b7", "messageHeadline": "Bump ruff in CI (#15533)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 326, "deletions": 199}}, {"node": {"oid": "c2221d919f1903f995f81c7889c85588df1849b2", "messageHeadline": "create a stream in DeviceStreamCollection for memory pattern (#15426)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 143, "deletions": 63}}, {"node": {"oid": "8fa65aba0eb4dfa475162b15d6cb6705b9969000", "messageHeadline": "enable training tests for csharp bindings (#15513)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 17, "deletions": 5}}, {"node": {"oid": "7ed3bfde513a7cad7ba8cc15e10ff82c0b030c94", "messageHeadline": "Fix FusedConv for ROCm (#15460)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 72, "deletions": 80}}, {"node": {"oid": "ac6ceffb2c13f1525fa2a52f61c3554cdea20fc0", "messageHeadline": "Force using fixed random seeds for flaky tests (#15515)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "5ebe700a9b7514bb6dfa951df34cffb4548196c0", "messageHeadline": "[QNN EP] Fix pool and conv op tests (#15504)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 63, "deletions": 19}}, {"node": {"oid": "fbe88fccbde56b551ec0fbe5d7616437ae9a07c8", "messageHeadline": "Exposing new TRT build options (#15089)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 363, "deletions": 20}}, {"node": {"oid": "4e1f75810c9aa115f922515a4577a9ac0fcf837c", "messageHeadline": "Add compilation cache in 2 Linux CPU pipelines and refactor the Linux\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 222, "deletions": 121}}, {"node": {"oid": "bf32dbbd9b61a5f8f2df85c3804165baee8a2e4b", "messageHeadline": "Share more constant initializers (#15461)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 417, "deletions": 69}}, {"node": {"oid": "f297bbb89b1696d3c85affa8f87d1488c5a154d0", "messageHeadline": "Fix an indent error in build.py (#15497)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "0fdd356abff9d5ae2bc259de2628cd5badd0b3c2", "messageHeadline": "[ROCm] Add hipBLASLt GEMM support to Tunable op. (#15351)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 254, "deletions": 1}}, {"node": {"oid": "fda0aa14c8b172d03a94dfe59663d17db064e325", "messageHeadline": "SkipLayerNorm fusion with different input and output type (#15500)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 183, "deletions": 68}}, {"node": {"oid": "d76cf374c48f1849193cc958e14f1851a723cb20", "messageHeadline": "Capture both ValueError and RuntimeError (#15503)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 12, "deletions": 8}}, {"node": {"oid": "56ad68120e9beddd7c1b7a159dab59e75b6242fe", "messageHeadline": "Add support to use sequence as input ids in decoder inputs to Beam Se\u2026", "author": {"name": "Akshay Sonawane", "email": "111780983+apsonawane@users.noreply.github.com", "user": {"login": "apsonawane"}}, "additions": 45, "deletions": 24}}, {"node": {"oid": "5bed8d028548e52e26886a860ad678db7316c017", "messageHeadline": "Disable XNNPack EP's tests in Windows CI pipeline (#15406)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 39, "deletions": 172}}, {"node": {"oid": "05ec22330f1d7195d88b57bf1a0e2e283bbd900f", "messageHeadline": "softmax perf improvement pr2 -  import softmax bw (#15199)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 134, "deletions": 23}}, {"node": {"oid": "67ac36101cd9f81c882e94582a4075521559582b", "messageHeadline": "disable BatchNormalizationGrad test (#15485)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a22dc65a81387f30979fcb5b421be2d4c36116f8", "messageHeadline": "Add a missing header to cuda_common.h (#15489)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "516c8e95fac59d50224630a2c95b0c54954efb00", "messageHeadline": "Optimize SCE loss compute (#15401)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1215, "deletions": 245}}, {"node": {"oid": "07b64d5275ecb5f2d1d0332842eb88c31139bf61", "messageHeadline": "Remove codecov from requirements-dev.txt (#15487)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "fd7f0c3cfcafebc56e0b11e100a1ec2cb4a723b5", "messageHeadline": "[DML EP] Use ORT node names in DML execution plans (#15411)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "e361e3f13825c5df286716ff8d32f8591c36eda3", "messageHeadline": "Fix bug in handling of variadics in function schema creation (#15409)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 36, "deletions": 10}}, {"node": {"oid": "e1e88522133950fe49a07656bf7027c416df6ecc", "messageHeadline": "[build/npm] dump ORT_COMMON_FROM from validation (#15475)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "3875c824d5c5ac9c0607ac3b56e6a0584c47eb61", "messageHeadline": "[build] fix default value of flag cmake_generator (#15471)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "041a0e274734d0ea2c0df232b75a79468e05c102", "messageHeadline": "[build] fix nuget linux in build_protoc_for_host() (#15472)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8cd5f3ad9cf2367db8f5fd4f69bee7caa9ed59f1", "messageHeadline": "[TensorRT EP] support TensorRT 8.6-EA (#15299)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 142, "deletions": 334}}, {"node": {"oid": "e3086b2ed84c844dec6220d6fad26e9425650fac", "messageHeadline": "Move DML CI Pipeline to A10 (#15468)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 11, "deletions": 9}}, {"node": {"oid": "00165540900f08b658f653d524f90fedadc21230", "messageHeadline": "[ROCm] disable composable_kernel and kernel explorer for MIGraphX CI \u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "d49a8de9b1db3486246e65700380e1eb8e7d581d", "messageHeadline": "[ROCm] add FP16 support for FusedConv Op (#15443)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 73, "deletions": 24}}, {"node": {"oid": "ce1eb6d6299b1b71a3b1f6b3647ec6aabfb6df94", "messageHeadline": "[ROCm] Add Tunable GroupNorm (#15298)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 886, "deletions": 383}}, {"node": {"oid": "db4fc12318575cecaa1ffa09e68fc1f596dff18d", "messageHeadline": "Add support for building the code on Windows ARM64 natively (#15371)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "9c42d5e31f70cf2e9f2f13a149b70bff2141511d", "messageHeadline": "[CoreML EP]Add broadcasting support for binary ops (#15187)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 113, "deletions": 17}}, {"node": {"oid": "0fbf715824ccb39b25ca4df754b37b01eb38e1ba", "messageHeadline": "[build] add script to validate generated NPM packages (#15453)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 183, "deletions": 8}}, {"node": {"oid": "ce3b4eabd35350824d6d4ebffaadfb2a31cb31ed", "messageHeadline": "Implement Optional Metadata support and C# test support (#15314)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 6563, "deletions": 1017}}, {"node": {"oid": "0497ac04326dd7b0a3fb4795b56dd35bd5868782", "messageHeadline": "Support additional op domains in op reduction script. (#15424)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "3be5bfe363ad602c6aa96c2d42df6028d55158d9", "messageHeadline": "[DML EP] Add MatMul + SoftMax fusion (#15240)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 272, "deletions": 3}}, {"node": {"oid": "7c927bb95c2eb0c3ac23908821ee9c0e3e1c24b6", "messageHeadline": "[DML EP] Add BiasSplitGelu (#15197)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 177, "deletions": 5}}, {"node": {"oid": "311f84d00c156b69b663fab63a35d3316262122a", "messageHeadline": "Fix one nuget packaging pipline error  (#15458)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "29c74d3c43923cafa9099cdd523fd5d64c814b7f", "messageHeadline": "softmax perf improvement pr1 - add more softmax related test (#15176)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 173, "deletions": 37}}, {"node": {"oid": "ef42fd09fb67af8fe321d01238a90fd1f5ac42e1", "messageHeadline": "google/mt5 optimization and fix (#15454)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 42, "deletions": 17}}, {"node": {"oid": "c5b6ee1a9904ff2df8e8ff5c78413cee57854545", "messageHeadline": "[DML EP] Add NhwcConv (#15194)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 129, "deletions": 58}}, {"node": {"oid": "9acbfc6a299e9a5a43037fd68b709ca3e291167d", "messageHeadline": "ROCm MHA (#15279)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 467, "deletions": 76}}, {"node": {"oid": "feafbc42637588aeef9a074bdabe3320f0579b70", "messageHeadline": "Refactor all Mac build steps (#15440)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 120, "deletions": 131}}, {"node": {"oid": "d175e87a1f2740a7b6a01aa350cc554a0ea06dbf", "messageHeadline": "Delete eager mode code and increase minimal required python version t\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 13732}}, {"node": {"oid": "4a676b011a71f5a1ddacdac566f5579f0e132976", "messageHeadline": "[DML EP] Add BiasAdd (#15211)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 124, "deletions": 2}}, {"node": {"oid": "ce9ad8c8bccf25d89dbd8983ed3d29609f3d6607", "messageHeadline": "For HLSL shader ops in the DirectML EP (STFT,DFT) FP16 ops should fal\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 46, "deletions": 2}}, {"node": {"oid": "6657df921254261103a1e3bd65f0a4cb48fccae5", "messageHeadline": "[CoreML EP] Add support for LeakyReLU activation layers (#15327)", "author": {"name": "Shukant Pal", "email": "SukantK2002@outlook.com", "user": {"login": "ShukantPal"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "0205b637564b3559cc1409fefabb29de5293596c", "messageHeadline": "[wasm] optimize default session options parsing (#15428)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 24, "deletions": 49}}, {"node": {"oid": "c8524d2dabdf062ef6c8c94040b07eba2e04f222", "messageHeadline": "Refactor web-ci pipeline and delete eager mode CI pipeline (#15416)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 46, "deletions": 282}}, {"node": {"oid": "9ef11f1c6a2b8ef3378e06af554f0361a00cbd6a", "messageHeadline": "[QNN EP] Qnn batchnorm Op support (#15222)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 316, "deletions": 7}}, {"node": {"oid": "0ea965c541adbbe5d3f1ef29741be7548ff64772", "messageHeadline": "clear cache stat. after building (#15439)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 15, "deletions": 10}}, {"node": {"oid": "6d126f8996362b9de09490d722ea50b2c35dbdcf", "messageHeadline": "Add FP16 support for Whisper model (#15427)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 55, "deletions": 14}}, {"node": {"oid": "34f22daf25b287e2439c414ae2d83d6bd34a4dcf", "messageHeadline": "Support T5 Beam Search with DecoderMaskedMHA (#15386)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 657, "deletions": 168}}, {"node": {"oid": "f77c8f486331d9fc1a699e2370a7dd185619790e", "messageHeadline": "Fix Npm packaging pipeline (#15425)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 13, "deletions": 10}}, {"node": {"oid": "56beac4b5b493e6b3ae3541e73643c0d78269cb9", "messageHeadline": "VIT model handling in the Benchmark.sh file (#15045)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 75, "deletions": 25}}, {"node": {"oid": "3c5d02a9ce8fde56a85aa1156d628379485d38bd", "messageHeadline": "Implement BatchNormGradient kernel for CPU EP (#7622)", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 186, "deletions": 12}}, {"node": {"oid": "5e2f46df2bd15364d8629115a1d496f2d8ccf574", "messageHeadline": "update deepspeed version 0.8.3 (#15415)", "author": {"name": "Rui Ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "666aff56a4a079cfece596d4c697020c66e3c94a", "messageHeadline": "Add workflow to update Objective-C docs. (#15413)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 54, "deletions": 19}}, {"node": {"oid": "8db86f2c52663f74483672d1b8c9ffc6fd4eea7f", "messageHeadline": "Use fixed version of Android NDK in binary size checks pipeline. (#15\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "9bb4e4bef4f472de29effe8e700c5c066815b42c", "messageHeadline": "Fix masm flags (#15417)", "author": {"name": "Yateng Hong", "email": "yatengh@microsoft.com", "user": {"login": "toothache"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "c294040bac0e34bd7ef0cb97424bace7998900e7", "messageHeadline": "[QNN EP] Support AveragePool operator (#15419)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 297, "deletions": 13}}, {"node": {"oid": "139f3df4d258eda944a325f032b63873c4f1a7e8", "messageHeadline": "Update binary size checks pipeline to use stages for separate checks.\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 174, "deletions": 150}}, {"node": {"oid": "8dce83a8180b934bfb4adbbabb3957504db960e2", "messageHeadline": "Fuse 'Add' operator into FP16 Conv (#15213)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 430, "deletions": 170}}, {"node": {"oid": "bb21031cbbc4ddb894648d40bf43eaa20f8fcd76", "messageHeadline": "[QNN EP]Fix issue in LeakyRelu Opbuilder for HTP backend. (#15356)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 120, "deletions": 14}}, {"node": {"oid": "16f5909f2d2ca1cd1330d4298f4f9fba844576ae", "messageHeadline": "Introduce shrunken gather operator (#15396)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 415, "deletions": 12}}, {"node": {"oid": "d31dd5935a94819c9959dc016c4921f22ddd7f9f", "messageHeadline": "[QNN EP] Support Resize's pytorch_half_pixel coordinate transformatio\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 553, "deletions": 34}}, {"node": {"oid": "03dd4e6da340a8967d30504365120c5aa7917062", "messageHeadline": "[QNN EP]fix bug in DlError (#15412)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "df11c8595529c4518e6127696664a0a203a55974", "messageHeadline": "Download protoc.exe from nuget when cross-compiling (#15395)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 72, "deletions": 110}}, {"node": {"oid": "65579021eea5deca390bc4c1dd487d763e0e2d44", "messageHeadline": "Remove UTF-8 BOM (#15026)", "author": {"name": "Yuriy Chernyshov", "email": "thegeorg@yandex-team.com", "user": {"login": "georgthegreat"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e5617617fc8ae6c6c7ccf76b71be488c9a8d4574", "messageHeadline": "Float to float label encoder (#15400)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 42, "deletions": 1}}, {"node": {"oid": "276c0a00e4a0630a0d8c219e5a487b801df0d2b9", "messageHeadline": "Reuse QDQConv for ConvTranspose to generate the QDQ model (#15385)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 170, "deletions": 1}}, {"node": {"oid": "2bd8e4a130bbe4e9a5cb63aec60b48f3b7346626", "messageHeadline": "Petermca/whisper dedup (#15365)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 16, "deletions": 5}}, {"node": {"oid": "dc1845a9c8e224e5709c1c3f7283625cea68eaa4", "messageHeadline": "Update mimalloc dependancy to the latest release (2.1.1) for Windows \u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "d0cca91cfb1b4737efb2f198429439f91d2c2ee7", "messageHeadline": "Fix token_id values for whisper export (#15362)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "55495cc8099a8fd5cd06c4f827da8703e1f69d8f", "messageHeadline": "Do not apply QuickGeluFusion if an intermediate tensor is a graph out\u2026", "author": {"name": "Deokhwan Kim", "email": "91861+dkim@users.noreply.github.com", "user": {"login": "dkim"}}, "additions": 69, "deletions": 2}}, {"node": {"oid": "026fb3ca1ee755fff724a9cc840a81cc4bdb8a04", "messageHeadline": "Fix compilation error when CUDNN_HOME is defined. (#15348)", "author": {"name": "Stephan Gocht", "email": "steohan@gobro.de", "user": {"login": "StephanGocht"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "0fbbb6a43e43fcbcfa176736d3601c65de92acd8", "messageHeadline": "WindowsAI build failing due to deprecated .NET5 SDK missing in build \u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 90, "deletions": 86}}, {"node": {"oid": "a5b4d2a8a783335c8a9fff42518d6b221960c28a", "messageHeadline": "XNNPack: allow users to choose whether enable CPU MEM arena or not (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "ca68ab612623521e9b10f425c261d5d6a8d60ddb", "messageHeadline": "Support decoder masked self attention for greedy sampling (#15319)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 18, "deletions": 11}}, {"node": {"oid": "71a4e7eb97e842e1dfc419d75ffef811f721e5ad", "messageHeadline": "Automatically enable tunable op usage for production models (#15156)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 224, "deletions": 74}}, {"node": {"oid": "2e52de265aa471ca83c88b50d19d57afd9052e30", "messageHeadline": "Upgrade remainding python to 3.11 removing 3.7 (#15321)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "6e1e808ec802c27c0defe63c931093d609ef47f9", "messageHeadline": "fix error unqualified call to 'std::move' (#15347)", "author": {"name": "Thuy Dao", "email": "59903692+iishiishii@users.noreply.github.com", "user": {"login": "iishiishii"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "962d8d2b191c3870e58d533983c61f75cf4bd593", "messageHeadline": "Add compilation cache in react native CI (#15329)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 352, "deletions": 314}}, {"node": {"oid": "a7d321e9dc8eded098d8267162d84f24b14a7dad", "messageHeadline": "String to string label encoder (#15379)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 42, "deletions": 0}}, {"node": {"oid": "ea6b32fea8ffe6b2de7027abbeed6258ff10d495", "messageHeadline": "Fix: Add def main() in onnxruntime_test.py (#15208)", "author": {"name": "Leso_KN", "email": "info@lesosoftware.com", "user": {"login": "leso-kn"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "ef11032c8956006e36c2ad26270ff28a41e9db89", "messageHeadline": "[java] Allows the creation and extraction of zero length tensors (#15\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 79, "deletions": 12}}, {"node": {"oid": "9191e0425902eeed92ee67de70bd0cf6c6cd89b6", "messageHeadline": "[DML EP] Add QuickGelu (#15220)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 139, "deletions": 1}}, {"node": {"oid": "a96e19abc4a49943ead82fde2290f219dee833b6", "messageHeadline": "Add type annotations to `onnxruntime_inference_collection.py` (#15364)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 31, "deletions": 17}}, {"node": {"oid": "764e489a007606234f213a95650b636ef59df951", "messageHeadline": "Adding FP16 Global Average Pool operator (#15324)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 243, "deletions": 25}}, {"node": {"oid": "a4e9a48345cdd307a91b0845d145174397beff39", "messageHeadline": "Reduce operators support for int64 type (#15358)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 175, "deletions": 24}}, {"node": {"oid": "9f5aa8e0218fc87a54eea2dedd98bcbb5ce16d31", "messageHeadline": "Add clog back to onnxruntime_EXTERNAL_LIBRARIES. (#15363)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "a0d8dbe28d7d201444bc833463e996db5dbd136a", "messageHeadline": "Register Resize op into nhwc schema for Qnn EP (#15373)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "4db10c93d131c8737bce4550676e8744d8a2d244", "messageHeadline": "[TensorRT EP] make --use_tensorrt_builtin_parser the default behavior\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "fe0db63dee7a0943606ee244a52ede3c135f7abd", "messageHeadline": "Upstream reshape of merging batch/sequence (#15023)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2607, "deletions": 595}}, {"node": {"oid": "6b755debbc440e4bf20af0ed18348d1a7a9b94c2", "messageHeadline": "Miscellaneous updates to training artifact generation (#15315)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 238, "deletions": 23}}, {"node": {"oid": "198994d01d4e226fd3a8816e8b2fad00c84a691c", "messageHeadline": "Register PytorchAtenDomain in RegisterOrtOpSchemas (#14567)", "author": {"name": "Nhat Nguyen", "email": "nhat-nguyen@users.noreply.github.com", "user": {"login": "nhat-nguyen"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "5294cd0c55eee7de6cff790258a00c73354244cb", "messageHeadline": "Print value errors in ort.InferenceSession to user (#15360)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "207c57219abe618bc82ee86ce856dc5abdbc1ad2", "messageHeadline": "Add support for full ViT optimization (#15289)", "author": {"name": "Anton Korablin", "email": "39565513+iAnt0n@users.noreply.github.com", "user": {"login": "iAnt0n"}}, "additions": 63, "deletions": 14}}, {"node": {"oid": "1c1d3865615a46ab45c1563e720e6a7ca7bb4760", "messageHeadline": "Adds int32_t and uint32_t clip kernels (#15306)", "author": {"name": "Aditya Goel", "email": "48102515+adityagoel4512@users.noreply.github.com", "user": {"login": "adityagoel4512"}}, "additions": 41, "deletions": 3}}, {"node": {"oid": "adb3d5dcb96d8e609f90a308c04948d2681b1155", "messageHeadline": "Allow constant folding nodes that have missing optional inputs (#15344)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 34, "deletions": 0}}, {"node": {"oid": "4400e80452946ec7841458973659763a83785264", "messageHeadline": "Allow `Path` objects for deserialization of ONNX models (#15307)", "author": {"name": "Severin Simmler", "email": "s.simmler@snapaddy.com", "user": {"login": "severinsimmler"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "af28754e6f8ed9a08c0b56d832641db8af58c5e2", "messageHeadline": "Update python package pipeline to support 3.11 (#15311)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 83, "deletions": 26}}, {"node": {"oid": "0412bffbb480bf1e8dad804a0d2a0c66b7b46719", "messageHeadline": "fix build bug when enabling DEBUG_GENERATION (#15338)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "1251964f96eb69082870f51e3391c0763d29f038", "messageHeadline": "Petermca/beamsearch whisper (#15339)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 386, "deletions": 58}}, {"node": {"oid": "b54ca9a041b7d2baf870bd7301625f676d915066", "messageHeadline": "Read the cache in main build if it's a (Intermediate)merge branch. (#\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "5baf5f506b691ad5e5052ecc670c3afa40c3be9d", "messageHeadline": "log level control + fix typos (#15302)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 95, "deletions": 49}}, {"node": {"oid": "f30e2d4387e996492d6c7390965e941dc1fe867a", "messageHeadline": "Whisper Export (#15247)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 1631, "deletions": 0}}, {"node": {"oid": "3cf3fa04672ca17ec1e18676254c09ab8c691f71", "messageHeadline": "Fix prefast warnings (#15340)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 29, "deletions": 27}}, {"node": {"oid": "dec11afb834a49a00c8b73049dcd7fec919fe735", "messageHeadline": "Fix a prefast warning (#15343)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "44027797b0592d9d96be4d437c3bd6c81a72918e", "messageHeadline": "[QNN EP] Gather support int64 indices input (#15317)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 210, "deletions": 28}}, {"node": {"oid": "85bb13345d13e0172b5ffa113a9544267f74bee6", "messageHeadline": "Rework some external targets to ease building with `-DFETCHCONTENT_FU\u2026", "author": {"name": "Matthieu Darbois", "email": "mayeut@users.noreply.github.com", "user": {"login": "mayeut"}}, "additions": 38, "deletions": 32}}, {"node": {"oid": "e4aae94f20b995fe4b80643499d46af211e38fc2", "messageHeadline": "Remove azure build to unblock PRs (#15336)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 0, "deletions": 20}}, {"node": {"oid": "fbfe92f66af3227b87b3f6a981577225e10abb6b", "messageHeadline": "DecoderMaskedMultiHeadAttention enhancement  (#15292)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 897, "deletions": 111}}, {"node": {"oid": "7ccdf9ad8c95c03df62f145af6343909c1ce77db", "messageHeadline": "User/sheilk/sequence fix (#15239)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 129, "deletions": 21}}, {"node": {"oid": "c06ab5e353889350b26f0e67e913d224efd1e2bb", "messageHeadline": "Optimize use of Eigen::DenseBase::select() for PRelu (#15287)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 26, "deletions": 16}}, {"node": {"oid": "fff75a301c12c369885c4aa08d719b9914111f9e", "messageHeadline": "ORT_Web - JS graph parsing update  (#15185)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 42, "deletions": 24}}, {"node": {"oid": "c68044cc4b29b6231967a7ca00b73e3e1b4ea747", "messageHeadline": "fix prefast warning for GenerationCudaDeviceHelper::ProcessLogits (#1\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "c08d6b42e8f457996a248a74222ead18f11f2a62", "messageHeadline": "Add tool to support packing mode for BERT model (#15283)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 598, "deletions": 58}}, {"node": {"oid": "027e231a8351d2abe4ac394b9b41d3b167c630a5", "messageHeadline": "Report unsupport reason during tuning (#15246)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "60cc082f0ae69a51d4f49c221be8a97dbb905d48", "messageHeadline": "[NNAPI] Minor fix (#15052)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 6, "deletions": 9}}, {"node": {"oid": "d80859f63d1634bd608a5937f418d3e3c69564aa", "messageHeadline": "[ROCm] fix python packaging pipeline and add python10  (#15282)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 47, "deletions": 7}}, {"node": {"oid": "e870089ca838f73cd06357a164a4d63b74daba08", "messageHeadline": "Refining the offline tooling for training artifact generation (#15212)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1229, "deletions": 754}}, {"node": {"oid": "818b94b4eaa9f894cdf489ce2cd5ee3536f67bcc", "messageHeadline": "Add owners for public facing API files (#15288)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "605c2f4b89afbda363856cf7a6982f42f178f12a", "messageHeadline": "Remove fp16 support from apple (#15270)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 22, "deletions": 14}}, {"node": {"oid": "4645726d7478e3013e781230df4eb48676dca7da", "messageHeadline": "fix for webgl lrn (#15236)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 4, "deletions": 15}}, {"node": {"oid": "9f942e1a3e6eae55653205b0acc048ed406e96af", "messageHeadline": "Graph transformer to ensure unique DQ nodes for QDQ node units (#15145)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 690, "deletions": 267}}, {"node": {"oid": "786f8b98f7a8ed6165212bed48c7c01fc86ea624", "messageHeadline": "Add a page in the documentation for every operator in onnxruntime (#1\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1110, "deletions": 43}}, {"node": {"oid": "dc61d3b5b6942d7a14b3d22d9273cc44a47f7f11", "messageHeadline": "Fix symbolic shape inference script on precision loss issue (#15215)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 80, "deletions": 12}}, {"node": {"oid": "c2dad6893b9c225e21514417afcde06a15a86fee", "messageHeadline": "use cudaStreamNonBlocking flag (#15258)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "75f6861cb8d138ef7265bc7967ae665551a0bbaa", "messageHeadline": "Skip DNNL's opset18 tests (#15275)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "6d464748ba7fed2275ecba3a7406298cabc93438", "messageHeadline": "Make internal nhwc schema registrations complete (#15278)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 46, "deletions": 13}}, {"node": {"oid": "c5f5e3ec5ec8f324df8f9b2e6bd6b0a4f0f31ef9", "messageHeadline": "Improve 2 cache tasks in one pipeline yaml (#15267)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 58, "deletions": 17}}, {"node": {"oid": "aab3c15585c87fbf3830c2595865b693fc68d261", "messageHeadline": "Add Compliation Cache in CoreML pipeline (#15259)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 65, "deletions": 10}}, {"node": {"oid": "2928fda49015f638b51da6559aaae9187e425d4a", "messageHeadline": "[web] disable browser test temporarily (#15280)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "15f7dca9fb2aa6d107ad1c8bf9de7c6fa16256f1", "messageHeadline": "Update protobuf to 3.21.x (#15245)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 178, "deletions": 57}}, {"node": {"oid": "5d1dbfb43254e281a583794f419ba7d78a6b5488", "messageHeadline": "Update ONNX test data (#15256)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4a0b86eba6395397719b33e08b1bead85f74375a", "messageHeadline": "Update the post-merge pipeline (#14965)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 130, "deletions": 176}}, {"node": {"oid": "fb1f03fdffe2b9198276fbcc6b84e7f6bb1585d1", "messageHeadline": "Increase the timeout value of win-wasm-ci.yml (#15257)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ecb89ed75219a8dacddf85d7a075f487a040044f", "messageHeadline": "[CANN] Multi-stream execution support for CANN EP. (#14058)", "author": {"name": "FFFrog", "email": "85333600+FFFrog@users.noreply.github.com", "user": {"login": "FFFrog"}}, "additions": 557, "deletions": 436}}, {"node": {"oid": "febc69e1b26e31b73fae3bb0b1b9501b727f0f69", "messageHeadline": "[QNN EP] Support Cast in HTP backend (#15234)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 272, "deletions": 4}}, {"node": {"oid": "a6279d4cfb51be98a5cc25dc642013e358fcd01f", "messageHeadline": "[ROCm] update Stable Diffusion benchmark to support ROCm EP (#15094)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 110, "deletions": 13}}, {"node": {"oid": "85948d6bc6df73b663a41ee1aad521b5fcab07d9", "messageHeadline": "Cjian/windows update python3.11 (#15243)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 56, "deletions": 20}}, {"node": {"oid": "659118f93914b3ca7bb6e0be5e58de0201ff3449", "messageHeadline": "Prefast warning fixes (#15175)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "f752bb9973bae6a8aefbf4a523a31268c22a3d55", "messageHeadline": "Update stable diffusion benchmark results: A100 and PyTorch 2.0 (#15195)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 48, "deletions": 24}}, {"node": {"oid": "710d095124cc7fc74570c5b9f5257ded0ec43fdb", "messageHeadline": "Refactor the constant `_ONE` in `orttraining_test_ortmodule_api.py` (\u2026", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 16, "deletions": 19}}, {"node": {"oid": "41ddcd30a1fa9233b4d5314f0e94c7abbc147d93", "messageHeadline": "Fp16 NHWC Max and Average Pooling (#15181)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 955, "deletions": 0}}, {"node": {"oid": "021e46179a649522c1f57236a23d5416e529961e", "messageHeadline": "[ROCm] refactor GroupNorm to set vecterize number as template paramet\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 53, "deletions": 98}}, {"node": {"oid": "938e2136c6c462448b731d61ee1fbeb8ba5282e5", "messageHeadline": "Enable pylint and numpy rules (#15218)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 120, "deletions": 118}}, {"node": {"oid": "62b2947ac1df5ae4777d939ee195362bf89db931", "messageHeadline": "[ROCm] remove python3.7 from python packaging pipeline (#15230)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 31}}, {"node": {"oid": "462c6043b5e34753d6c38b2dc5e3b0c774ea4442", "messageHeadline": "Remove Win8 support (#15219)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 16, "deletions": 106}}, {"node": {"oid": "eb8f6c7c52fa7ffd3e9b6865a7646f8f243a124b", "messageHeadline": "Transpose optimizer enhancements (#15117)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 766, "deletions": 181}}, {"node": {"oid": "792d4111355bc20f8d86d7847693688c0645db81", "messageHeadline": "Update python 3.11 and remove 3.7 for Linux (#15214)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 68, "deletions": 228}}, {"node": {"oid": "ea40dc3ad61004078531133c78e385d7a4cbade8", "messageHeadline": "Update build.py to disallow running as root user by default. (#15164)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 72, "deletions": 48}}, {"node": {"oid": "3064fa7611b55154155a60b1af8bb596c8381780", "messageHeadline": "Fix C API docs error (#15216)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 5, "deletions": 7}}, {"node": {"oid": "527e006124fe8542f5dd20d128fcc2ed7f39fd4f", "messageHeadline": "Update mlas (#15228)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "063ee8d50468d8b0032e80830f0e887d53038f3d", "messageHeadline": "Fixed some warnings that were treated as errors when compiling with D\u2026", "author": {"name": "Bengt Gustafsson", "email": "bengt.gustafsson@beamways.com", "user": {"login": "BengtGustafsson"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "63cc1bb26a7be4093a4cc9e004a8eee3d29e53bd", "messageHeadline": "Move  Linux CPU pipelines to an AMD CPU pool which is cheaper (#15144)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 75, "deletions": 72}}, {"node": {"oid": "67a6022c03dd948400cec5877043245a23c41a9e", "messageHeadline": "[DML EP] Add GroupNorm (#15189)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 241, "deletions": 2}}, {"node": {"oid": "2e566206110d2fc2a96e45a785d94b37bac9daf2", "messageHeadline": "Add file and line info in CudaCall and RocmCall macros (#15148)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 110, "deletions": 113}}, {"node": {"oid": "ffcfb1ec98d6c5b21abf6172ad71793093b040a8", "messageHeadline": "Remove protobuf submodule (#15190)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 185, "deletions": 24}}, {"node": {"oid": "e754edaecfca25ddd8dbf5a32ada1750dad6bf44", "messageHeadline": "Run rustfmt in CI (#15217)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 21, "deletions": 0}}, {"node": {"oid": "b10aaf4e9c6acbfbeefbafeb988319610a9ab840", "messageHeadline": "Fix error message when running NhwcConv with a bad weight channel cou\u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "d182d34f1df21ff9c59a905e1c25d988594a1ddb", "messageHeadline": "pause caching docker image in pipeline cache in Linux Aten Pipeline (\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d24b630fc3138ce14da0d40fdf3288dea1bcaa60", "messageHeadline": "[QNN EP] Support reduce ops with axes as initializer input  (#15126)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1180, "deletions": 286}}, {"node": {"oid": "d3565779c31d6bdebde4b05579e7209594c712df", "messageHeadline": "Allow bert_perf_test.py to load/save tuning results (#15096)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 47, "deletions": 3}}, {"node": {"oid": "93e69027903252ef6d0d2e2e3dd34be0bbd789fb", "messageHeadline": "resolve undefined symbol: rocblas_create_handle (#15204)", "author": {"name": "Chris Austen", "email": "causten@users.noreply.github.com", "user": {"login": "causten"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "750747d8c9f1a96bda182b9a8810f3f6a75d6bdc", "messageHeadline": "Cjian/multi stage packaging pipeline (#14993)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 2526, "deletions": 2340}}, {"node": {"oid": "5a2e43bdd5f03450015c7b3e33eb25264a8f73c0", "messageHeadline": "[QNN EP] Improve Slice to support opset 9 (#15186)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 35, "deletions": 12}}, {"node": {"oid": "d834ec895af8c2dd6e718c21b575697c01b8ec4f", "messageHeadline": "Adopt linrtunner as the linting tool - take 2 (#15085)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 3974, "deletions": 4099}}, {"node": {"oid": "2de15c5d503cf6e4d679a2b8fca7135adc9f5ae8", "messageHeadline": "Re-work OrtApi struct to satisfy C++20 compilers (#15183)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 4}}, {"node": {"oid": "dc87691000ceca1b42e3c46b1ce9066a739505ed", "messageHeadline": "Enable DML graph fusion independently of graph optimization level (#1\u2026", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 31, "deletions": 2}}, {"node": {"oid": "56bccac35d10042fc3c6a37e10335d134602206c", "messageHeadline": "[ROCm] update bert-L convergence reference file to fix CI (#15200)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "7eb6dbe7d862117b865cbf51a0329db0dae37779", "messageHeadline": "[ROCm] Add compute type for Skiplayernorm to fix ROCm CI (#15192)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 157, "deletions": 176}}, {"node": {"oid": "3a4c895765c7a917279f9ace442ae9f5119258ea", "messageHeadline": "[DML EP] Add support for SkipLayerNorm's fourth output (#15160)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 70, "deletions": 9}}, {"node": {"oid": "0402f930f22e8c7a7e1b4f8f807aba157e2a84b5", "messageHeadline": "exclude decoder files in hipify.cmake (#15188)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 5, "deletions": 8}}, {"node": {"oid": "5c5c345abc9d4afe3f87b0ce4db1272ed519f2df", "messageHeadline": "Add smoking tests for all CPU Packages. (#15153)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 100, "deletions": 56}}, {"node": {"oid": "338e6672dd5f9a42f94bde96ec2fe970f142432f", "messageHeadline": "use build.sourceversion in cache image key (#15019)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 9}}, {"node": {"oid": "ea245c94e78ef7fe6aa2ab02f2bba7e961a12388", "messageHeadline": "Add constant folding for simple QDQ Node Units (#15138)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 106, "deletions": 20}}, {"node": {"oid": "0200995058a16b9dcb5def5a9c047d9cd250ca69", "messageHeadline": "Bump webpack from 5.75.0 to 5.76.0 in /js (#15159)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "28f64066de828decef47709f9030f36976ec71e4", "messageHeadline": "Auto deploy API docs (#15088)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 204, "deletions": 190}}, {"node": {"oid": "44ba23e0f52d3aca3cc9b194497664a96752992b", "messageHeadline": "Rename DecoderMaskedMHA to DecoderMaskedSelfAttn (#15166)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 155, "deletions": 164}}, {"node": {"oid": "2ee822d4831d2043e3d588d601264ed96cd72bf3", "messageHeadline": "Extend memory efficient attention coverage in Attention/MHA cuda op (\u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 382, "deletions": 999}}, {"node": {"oid": "7033346605e0e6bf619ac6049356d478f97e1237", "messageHeadline": "Support mask_filter_value attribute in DecoderMaskedMultiheadAttentio\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "910fc09de2af277dc54e01f99de3b6e4fc11c746", "messageHeadline": "Using standard layernorm cuda kernel for skiplayernorm. (#15076)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 153, "deletions": 332}}, {"node": {"oid": "88a66a289bfc5cea3177736f11d80b3425e18b72", "messageHeadline": "Fix prune_graph and gpt attention fusion scripts (#15147)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 33, "deletions": 25}}, {"node": {"oid": "b82f94ac2e5306ddf525a540f4011ccff792b699", "messageHeadline": "Update labeler.yml for web (#15142)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "559a21c7c30a82ca07a17409ea76d5793b0e942a", "messageHeadline": "Fixing CUDA12 build (#15135)", "author": {"name": "Boris Fomitchev", "email": "borisfom@users.noreply.github.com", "user": {"login": "borisfom"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "efa12626141db4c249b0bab87b4e0595b8b7172a", "messageHeadline": "Handle unused function inputs (#15130)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 30, "deletions": 1}}, {"node": {"oid": "896ab94780f1e934bf10199d6d2ec003d99930d9", "messageHeadline": "Remove root in run_python_dockerbuild.sh (#15169)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "1d322855363d0e9c5aa6e8a43d76c2d5718ad7bf", "messageHeadline": "Statistics tool for ORTModule convergence parity (#15020)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 972, "deletions": 48}}, {"node": {"oid": "039ca1082202e38179ad2db0b076b87ebee1d4fe", "messageHeadline": "Move offline_tuning.py, so that the utility will be package with whl \u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "11dac121b2d02226949a24be5dfe139c30daca33", "messageHeadline": "Fix test_custom_op_get_const_input inference test on Android CI (#15132)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "dccbe9d4925898cdde1c38d87a97a6ef53b02deb", "messageHeadline": "exclude packed_attention* from rocm (#15161)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "fc3a2a37718f89a9eb3f200bbcf15c76da3d7632", "messageHeadline": "[CoreML EP] Add Pad op support  (#14946)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 377, "deletions": 17}}, {"node": {"oid": "7bec80d92a1e654b23bc7b459a2ec6e95a79c22d", "messageHeadline": "Fix reference count for autograd.Function (#15121)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 209, "deletions": 72}}, {"node": {"oid": "f972d21e81a5b1bcc5c55f39ed0b7cb3b4b08bde", "messageHeadline": "[js] upgrade dependencies and enable strict mode (#14930)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 8874, "deletions": 17746}}, {"node": {"oid": "71b67ec1e2ad74b5eb1875d44ed329746aca104f", "messageHeadline": "Refactor ke register to be decentralized (#15036)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 84, "deletions": 294}}, {"node": {"oid": "0086f7590d2ecf5649e9c9ff22db794dda186695", "messageHeadline": "LSTM and LSTM gradient implementation for training (#15034)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 3005, "deletions": 18}}, {"node": {"oid": "126e7bf15fa4af8621814b82a3f7bd0d786f0239", "messageHeadline": "[AMX] add assembler check (#15055)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "3e2d453b641cff31fb01954ff611a7ca1f1531dc", "messageHeadline": "Supports model > 2GB in fp16 conversion with onnx shape inference (#1\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 52, "deletions": 13}}, {"node": {"oid": "c7ced7a5e9c5ba1245b03f84edd537f45432b2e9", "messageHeadline": "Add PackedAttention for packing mode (#14858)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1915, "deletions": 12}}, {"node": {"oid": "ef76b3aeb849c5ea008538316b39294b3c3fb277", "messageHeadline": "Transformers tool - update readme to link to docs page (#14964)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 17, "deletions": 164}}, {"node": {"oid": "34175f0b7cd949b6b719633f68fb223fa129fee5", "messageHeadline": "FP16 conv (#15062)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1924, "deletions": 0}}, {"node": {"oid": "a3570eb5bf98d55a85130545bc598856d0f055ae", "messageHeadline": "Add mac packages smoking test (#15122)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 64, "deletions": 4}}, {"node": {"oid": "abb2418c0281ed46f44ee8434605f2a58cf54095", "messageHeadline": "TensorRT EP - timing cache [patch] (#15113)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "0ace27fdf785ccb790ee407bd4dc3c218017bf78", "messageHeadline": "Disable unit tests for decoder masked multihead attention on CC 5.2 o\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "226a691e052a53cd3b7aa52cdd9f061469f0e13b", "messageHeadline": "one prefast warning fix (#14912)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bdd7bd084c8b3b82ca6641ebbd104b66481885ec", "messageHeadline": "Remove the use of eval in test code (#15097)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 122, "deletions": 214}}, {"node": {"oid": "c964da7ea2fcd6a231627365eaf17ee7be5bf32c", "messageHeadline": "FasterTransformer model wrapper using custom op (#15013)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 332, "deletions": 6}}, {"node": {"oid": "32a4eebc178ea8863019d3124d6ece73194a5489", "messageHeadline": "[ROCm] add rocm5.4.2 to python package pipeline (#15081)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "e42f7487dfe5b986a5a945a99ce0da99404e370c", "messageHeadline": "Add logging APIs for custom operators (#14416)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 929, "deletions": 20}}, {"node": {"oid": "2023836c2f97dd6fe417e0b6f2295efdd3346fb8", "messageHeadline": "Fix CUDA tests for Ampere cards, and bump layernorm tests opset versi\u2026", "author": {"name": "Kevin Chen", "email": "45886021+kevinch-nv@users.noreply.github.com", "user": {"login": "kevinch-nv"}}, "additions": 17, "deletions": 12}}, {"node": {"oid": "98ab4a62d647a95011667a3704ff46e7bc2d2920", "messageHeadline": "Fix ROCm 5.2.3 pipeline (#15073)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 41, "deletions": 22}}, {"node": {"oid": "1ccb79476cf960f360e3d1dd809971b169a81cd1", "messageHeadline": "Fix training gpu ci related to pl upgrade (#15092)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5ac3a37be5d5bd25d167c2127b9c85f4b53668e7", "messageHeadline": "Add float16 Tanh support (#15048)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 229, "deletions": 17}}, {"node": {"oid": "4a8cd4256a44260d925a9f88d24469248fadb6fd", "messageHeadline": "fix miopen new API cannot be supported by ROCm5.2.3 (#15077)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 34, "deletions": 0}}, {"node": {"oid": "55174bb2e92c3ca77a03843e15420b05010465fc", "messageHeadline": "Fix Nhwcconv with asymmetric padding (#15050)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 77, "deletions": 19}}, {"node": {"oid": "6a6513f9c0541e06ebede87cbf84ada0d63f1394", "messageHeadline": "Bump @sideway/formula from 3.0.0 to 3.0.1 in /js/react_native (#15028)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "3059a73b38a3737d86cf63047332503352c42af9", "messageHeadline": "Bump @sideway/formula from 3.0.0 to 3.0.1 in /js/react_native/e2e (#1\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 126, "deletions": 86}}, {"node": {"oid": "6600fd792a9a1ffd402ef7021fa7bdc119cb9732", "messageHeadline": "Fix CPU memory leak due to external weights not getting memory unmapp\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "1e7849c2c8a71d2d1a771bae05fce1ab6e573360", "messageHeadline": "Add compilation cache in iOS pipeline (#15070)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 39, "deletions": 0}}, {"node": {"oid": "25e537770fc637e0f5891f38ef494aa720632a6b", "messageHeadline": "[CUDA] Fix Alignment of SkipLayerNorm Vectorized Kernel (#15054)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 73, "deletions": 76}}, {"node": {"oid": "a5ab88247b6733f664fa469f9c09692aae5e8e1a", "messageHeadline": "ROCm Flash Attention (#14838)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 1304, "deletions": 64}}, {"node": {"oid": "881f3f6be3caec99d994bf9bde95b84ed5792195", "messageHeadline": "[Fix] Error in Linux_Packaging_combined_GPU of nuget packaing pipelin\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 26, "deletions": 1}}, {"node": {"oid": "ed7ab1660dc59a91d33ccf0cc71da2a672469ddd", "messageHeadline": "[CUDA] Add option to use DecoderMaskedMultiheadAttention in BeamSearc\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 645, "deletions": 174}}, {"node": {"oid": "da084b0fc163f131c117ceffe654adef992d6c63", "messageHeadline": "check axis range for LayerNorm (#14845)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 13, "deletions": 6}}, {"node": {"oid": "5213546e625a53106b6885b7a63ba4ac8614e331", "messageHeadline": "Change how to find npm (#15001)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 29}}, {"node": {"oid": "32533dd1c290aa6025c584a705d3fa00fde1cd0a", "messageHeadline": "fix", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "cc15ceef4e3c1f6c49942cf72ccd67f32ffd770c", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/builders/model_\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "6bdb03281a5f21cc07633627d4db1b783b4289e0", "messageHeadline": "clean comments", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 55, "deletions": 40}}, {"node": {"oid": "762ea2402e7e75b9e2443f3f87cccd6aa1bd6a7e", "messageHeadline": "fix", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 17, "deletions": 11}}, {"node": {"oid": "8db28d9139050654701b7ce2eaf36c633df2cd6a", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_api_helper.h", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8d0096132123454efab6dabb352934e55e613483", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_api_helpe\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c10462b5f505985643b25221501366513d4d644b", "messageHeadline": "ORT_UNUSED_PARAMETER", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "92fabf57ea194a0b91a57afcd0fe106619aba2b2", "messageHeadline": "comments", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 53, "deletions": 25}}, {"node": {"oid": "cd3173d53109582996e5728da58b43bcdd95493a", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_api_helpe\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dad772ef0966b30024fcdad204416fdf3624d5a7", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_api_helpe\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8aeed1e87d2c36609ee075a193f9cfd89c826f51", "messageHeadline": "amend", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 4}}, {"node": {"oid": "5fe61c53a3ae04e4dca7331355807806485b1b9e", "messageHeadline": "comments", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 34, "deletions": 37}}, {"node": {"oid": "d236085845e439e0489768eb22be25a217620716", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/model.cc", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d490908836f7317faa87c3c0856ff4aa1a744c22", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_execution\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a8ed956fa7d345a597c404d8c3bf2c4844b72592", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_execution\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "de5e58c07755bd81f191145425a713a3f482b233", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_execution\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d70b5b38e21de5d48283cf1373fbab2d63a887a7", "messageHeadline": "amend", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fdc970a40dead39ff766e3a5b1a59430071eb808", "messageHeadline": "rename constant var", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "18015f0f556ba790bc49bb77eee4f29c2b293745", "messageHeadline": "use span", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 65, "deletions": 60}}, {"node": {"oid": "4ca84ac30357690a0f23b0cef0c529d27aadd813", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/model.h", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "28a4cfeeef64e3729d5fbd089f9555fd3a91a607", "messageHeadline": "refactor", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 78, "deletions": 67}}, {"node": {"oid": "adfd38edb93dde5464747427d6cc6347b6fc1904", "messageHeadline": "Fail early when getting device", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 44, "deletions": 31}}, {"node": {"oid": "82ae1381439af31e8bb0cb029b8dafac14c78ced", "messageHeadline": "address comments", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 130, "deletions": 123}}, {"node": {"oid": "36aac0036bb4d891ae033fb228bece2f589102bd", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_api_helpe\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "760a2b99d0ea8ee9defdd02711687e392f391211", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_api_helper.h", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "7c6fc31b6584bd8f5a53f87db2245aba8f9dc7d3", "messageHeadline": "Update onnxruntime/core/providers/nnapi/nnapi_builtin/nnapi_api_helpe\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "adf990f30f8dfc716d3a401e7104e4cf0da070d0", "messageHeadline": "more docs", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 72, "deletions": 38}}, {"node": {"oid": "028c2372fac9e270aff6f7e1f49345d067e92c71", "messageHeadline": "remove disable_cpu_soft temporarily", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 23, "deletions": 44}}, {"node": {"oid": "30151323dad4205843ae836c63448e171d165be7", "messageHeadline": "move init_device to ctor", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 13, "deletions": 8}}, {"node": {"oid": "d135bb7c0c59e7b76e54b65e13eea2c3aed711af", "messageHeadline": "Update nnapi_execution_provider.cc", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "781e72f6634360550ff37a79f080c9d064c20f4c", "messageHeadline": "Update nnapi_api_helper.cc", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8383a54f9db71798114e88b5a39865e80ab2cb52", "messageHeadline": "Update include/onnxruntime/core/providers/nnapi/nnapi_provider_factory.h", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3873a55bd346d2993e0c4cfa2df665e9e0f621cd", "messageHeadline": "[NNAPI] fix feature_level query", "author": {"name": "wejoncy", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 286, "deletions": 111}}, {"node": {"oid": "6891ab5bac78b0f36c1f18731d4d6f2d92452dc1", "messageHeadline": "fix_macos (#15018)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "bdfdebfca737b235525a05ee6b574ab2e88a6aa6", "messageHeadline": "Fix ReduceSum in attention fusion (#15047)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 29, "deletions": 7}}, {"node": {"oid": "c70838cbbbbfbf3f3375984004162c1dc6a8a437", "messageHeadline": "[ROCm] add Conv, NhwcConv benchmark to microbench (#15017)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 134, "deletions": 0}}, {"node": {"oid": "f096f6167bf452c06371d14ffd23e271089eaf88", "messageHeadline": "Remove python37 and cuda37 packages in orttraing (#15041)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 11}}, {"node": {"oid": "a8680ff1885c7752068a207681e1777138e68a91", "messageHeadline": "Add float16 Sigmoid support (#14910)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 256, "deletions": 67}}, {"node": {"oid": "0bb7390d11c186414f35647da979bef22413235d", "messageHeadline": "fix prefast warnings (#15033)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 46, "deletions": 21}}, {"node": {"oid": "db4e664f7c5a7ee67e2d8a64b0d6d5e4d481376d", "messageHeadline": "Re-enable react native e2e android unit test for CI and upgrade targe\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 32, "deletions": 32}}, {"node": {"oid": "8b09702b88a2ed2258c5e5a5f63a5b539ac93f43", "messageHeadline": "Enable parallel computation in Clip ops (#14925)", "author": {"name": "Alex Kogan", "email": "82225080+sakogan@users.noreply.github.com", "user": {"login": "sakogan"}}, "additions": 49, "deletions": 11}}, {"node": {"oid": "2ff7f3e93a593329531347479781b081883013d8", "messageHeadline": "[ROCm] support optimized Stable Diffusion model  (#14980)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 931, "deletions": 160}}, {"node": {"oid": "ff2850029bcacf4ce4d5ec160b4689002b6d103b", "messageHeadline": "[ROCm] refact SkipLayernorm long if-elseif statements (#14795)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 28, "deletions": 59}}, {"node": {"oid": "0fa00429d561fcdea5330ebe817c69d52b2ad6fb", "messageHeadline": "[T5 optimization] script fusions and fixes (#14967)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 731, "deletions": 146}}, {"node": {"oid": "59dfcfdce712a5d527c7b986fc41a8e4e14c702a", "messageHeadline": "Fix typos in sources: operater, tranform, neccessary, trainig (#14907)", "author": {"name": "Christian Veenhuis", "email": "veenhuis@gmail.com", "user": {"login": "ChVeen"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "538d64891ac8e43c1faf7846635c3a1bf7b6b6c5", "messageHeadline": "[t5 optimization] kernel changes to t5 (#14928)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1345, "deletions": 152}}, {"node": {"oid": "b34e570ad0df3af3b2f3d9356b4aaf2f2729e1bf", "messageHeadline": "Enable LeakyRelu latest and refactor fast_gelu_fusion to enable the s\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 31, "deletions": 19}}, {"node": {"oid": "a5d814008c66c5efe50cd7153fc06a0075827078", "messageHeadline": "Fix API docs deploy so that a PR is not required (#15011)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 18, "deletions": 16}}, {"node": {"oid": "44dda08b51e98bd86f6eafd09b7d00c3680c78a8", "messageHeadline": "Renaming files (#15015)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 4, "deletions": 7}}, {"node": {"oid": "c55f3476897d8afb2ac89e04b990e31c46dfbbf0", "messageHeadline": "[ROCm] change miopen_conv_use_max_workspace=true (#14982)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "448e989df894c22fb384ca2d29433bf503c6f70b", "messageHeadline": "Op slicing upstream refactor (#14832)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 797, "deletions": 582}}, {"node": {"oid": "cce9e0eaad27399412fb8cdf3de5aa5769535444", "messageHeadline": "Add float32 hardsigmoid tests (#14948)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 80, "deletions": 28}}, {"node": {"oid": "930e009567f9b71505c5c4e99b13c4ed947707a0", "messageHeadline": "[WIP] Update call to GetFunction (#14949)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 31, "deletions": 3}}, {"node": {"oid": "ca315b9148274b8bfcd6721f147658d230fa280c", "messageHeadline": "Use ADO cache to cache docker image instead of ACR (#14496)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 86, "deletions": 4}}, {"node": {"oid": "79501899206524e6913bf3a375125c07c8f4a247", "messageHeadline": "[CUDA] Optimize Perf for AtomicAdd of Half Type (#14992)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 64, "deletions": 9}}, {"node": {"oid": "a8ad0edbeb45a1733d5b062acc13c6b3ad08731b", "messageHeadline": "BUG FIX: the if...else in  telemetry-steps.yml does not really work (\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 27, "deletions": 12}}, {"node": {"oid": "d8ddd25272c2946dd4a1bcbab5b9c05690887efb", "messageHeadline": "Add InstanceNormalization operator to QNN EP (#14867)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 782, "deletions": 9}}, {"node": {"oid": "a5c436e14874cea45dea850f4a3855d28be21626", "messageHeadline": "Fix prefast warnings (#14975)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "0d7855ea5a542de153eadd0a6b37d42a6f55fdcd", "messageHeadline": "Re-work global objects dependancies in pybind layer. (#14941)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 256, "deletions": 171}}, {"node": {"oid": "e2febe87f631d0d5969b7171090879f1325ad2bc", "messageHeadline": "[QNN EP] Update QNN SDK to 2.8 (#14978)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 62, "deletions": 35}}, {"node": {"oid": "bd142bfb04db00253265203fb0af2988e751d9a0", "messageHeadline": "Gradle clean up (#14973)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 14, "deletions": 54}}, {"node": {"oid": "748758c135b1ce4a909cffb65b950656f2019f4d", "messageHeadline": "Address issue with uninitialized variable (#14988)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "ad4db1269972f92fdba932bb5770943291be3ca5", "messageHeadline": "TensorRT EP -  timing cache (#14767)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 303, "deletions": 18}}, {"node": {"oid": "acbb7ad45335aa1721cad0180e94d626a0d45cb6", "messageHeadline": "enable cache in orttraining-mac-ci (#14979)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "1187d4ade62b8a0a98e7e0eb25e53d73959baf01", "messageHeadline": "[wasm] extend build timeout for static lib (#14952)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "79d47c153093c64a6099c0c307ac3df36931591a", "messageHeadline": "Enable sorting of initializers (#14631)", "author": {"name": "Preetha Veeramalai", "email": "preetha.veeramalai@intel.com", "user": {"login": "preetha-intel"}}, "additions": 13, "deletions": 5}}, {"node": {"oid": "b4fe98ac2e256fdb7bc2f40d9796ceb802883911", "messageHeadline": "Update to MacOS-12 (#14924)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 21, "deletions": 20}}, {"node": {"oid": "51b67fa15c703ea460aeef096745c3d591658532", "messageHeadline": "Make ROCm Attention biased+masked and biased+nomask scaling logic con\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "f83923d5dfbbe5e5b6fe890351d0c5b02c385c6a", "messageHeadline": "fix rocBLAS extensions API issue; add batched- and strided_batched- c\u2026", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 277, "deletions": 7}}, {"node": {"oid": "bf2cc808a180da4a02147271a4691fcbd83055bb", "messageHeadline": "[ROCm] SkipLayerNorm: add more configs for block size; loosen constra\u2026", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 45, "deletions": 13}}, {"node": {"oid": "d55ae490e19421a22ed17243bd2478a39ea72a56", "messageHeadline": "detach patch manylinux from get_docker_image (#14958)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 77, "deletions": 28}}, {"node": {"oid": "80e25ad6ac5e0e141a2c6f2107a25b78c0f9b690", "messageHeadline": "fix cg issue (#14372)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "3c4efd2e775fe60854dc08ab25c0396544fcdc51", "messageHeadline": "[js/common] allows polyfill for bigint (#14921)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 23, "deletions": 11}}, {"node": {"oid": "8844474083071e15262597e0430f929bd225b9a2", "messageHeadline": "[js] remove 'npm bin' (#14943)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 37, "deletions": 53}}, {"node": {"oid": "d8d96f0788278f1ce7d6dce3ba8528bc26521e8a", "messageHeadline": "Fix a build issue (#14944)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "c46c7ccba5b405e62d2d53be5bcda1a5e4851cfa", "messageHeadline": "Update Gradle version (#14862)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 435, "deletions": 119}}, {"node": {"oid": "d9436407b6d6c975179a299408da543e3c65e0f2", "messageHeadline": "Use safe allocator for JNI code (#13999)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 57, "deletions": 34}}, {"node": {"oid": "47f00b5d49ece70d5905f32a69147812a803f080", "messageHeadline": "[Java] Initial on device training support (#14027)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 2297, "deletions": 32}}, {"node": {"oid": "f14ab63c19582705464fe794debaf15178833c50", "messageHeadline": "fix prefast warnings (#14931)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "112a4d215acd5261c730103168d167de3a5c1f68", "messageHeadline": "[CUDA] Support decoding multihead self-attention implementation (#14848)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2978, "deletions": 40}}, {"node": {"oid": "c696392f0c709111ddf44b62024bfbd28c6651e8", "messageHeadline": "Support external output tensors for DORT (#14516)", "author": {"name": "Kyushick Lee", "email": "kyule@microsoft.com", "user": {"login": "kyule7"}}, "additions": 89, "deletions": 23}}, {"node": {"oid": "2ef25a22007f894e811553ab954b588b93897d56", "messageHeadline": "Update CODEOWNERS file.", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "5b3f79a11a253a524e2a78edcd7a059db7940bb4", "messageHeadline": "Add gradle wrapper validation workflow.", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "f71ac9859e98ab0871eb0b112ce57a12a846e10d", "messageHeadline": "Update acpt image in the training pipeline (#14855)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5d8ce817cb5651b8ff579ef91e8dcfb6cf8583ed", "messageHeadline": "Fix simplified layer norm fusion for training (#14866)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 26, "deletions": 10}}, {"node": {"oid": "65f1f840f6132663dd1645f2b9681237fe02c1e5", "messageHeadline": "[DML EP] Fix Attention regression caused by removing transposes (#14908)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 79, "deletions": 32}}, {"node": {"oid": "6b604521a6755ce2fdf18ac3725c480e55535f39", "messageHeadline": "Fix tree implementation when left, right node have lower index (#14839)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 148, "deletions": 55}}, {"node": {"oid": "66101c02a2ca0b0b7942d60b22f2899367556544", "messageHeadline": "Implement AllToAll collective op", "author": {"name": "Hitesh Shah", "email": "hiteshshah@microsoft.com", "user": {"login": "shaahji"}}, "additions": 117, "deletions": 7}}, {"node": {"oid": "150043f74fc9401c128074a9313e9be6c70475ec", "messageHeadline": "Adds a Java accessor for GetVersionString (#14876)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 65, "deletions": 5}}, {"node": {"oid": "5930e7e22fea92695b577731dd0ac2bc44afea26", "messageHeadline": "Introduce RemovableAttributes (#14868)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 119, "deletions": 4}}, {"node": {"oid": "be1416d03250ee143c4058dfbe040cdef023e90a", "messageHeadline": "prefast C26451 (#14933)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3e08a67dd6c98333a97f63bfca581f5abba94ef6", "messageHeadline": "Add Linux ARM64 CI pipeline (#14904)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 211, "deletions": 177}}, {"node": {"oid": "d45b47945c85062aa33477f3b794bc84501447dc", "messageHeadline": "Linux QNN Pipeline: fix build error reporting (#14922)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "f88b97ede272e54d67e7f4da165359df4b534a4a", "messageHeadline": "Cast to int32_t->size_t to avoid prefast overflow warning (#14902)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6c8538f0862f213ba8df3ffde67e81024a7a810f", "messageHeadline": "Fix prefast warning (#14895)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c1155b70c5c46e9e879fbbdd34a6aa2624951e7d", "messageHeadline": "Remove 37 and 50 from CUDA compute archs (#14874)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 15}}, {"node": {"oid": "289f7dbcdd70c201e6fd4fe0aa184b82c04945e9", "messageHeadline": "enable pybind for qnn ep (#14897)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 21, "deletions": 0}}, {"node": {"oid": "f6c81d8aca6c32a0866e5063fb710669023c33d5", "messageHeadline": "Introduce padding inspector in ORTModule (#14652)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 334, "deletions": 3}}, {"node": {"oid": "8c454a76e0040f479218b7c2213838d40118a6f1", "messageHeadline": "Check Mac silicon package name  (#14898)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "a997bb46b6a3cd3f1c631d5d15a9d488ca22dff8", "messageHeadline": "Refactor rocm attention (#14688)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 628, "deletions": 470}}, {"node": {"oid": "f3b6664384aeed424692956eee0ff8c5a666c7ef", "messageHeadline": "Remove Python 3.7 from the python packaging pipeline (#14887)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 35}}, {"node": {"oid": "c49f250a14bf58666604ac417dff41aca1dce8e5", "messageHeadline": "Del ort_model._modules to foward its accessing to torch_model._module\u2026", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "8d87fdcfa19a6b5932c88137b9cd6a43a658c7f3", "messageHeadline": "Add GetVersionSting API for C++, C# and Python (#14873)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 58, "deletions": 3}}, {"node": {"oid": "6e2ca15140d0bb721dfe0184c070bd1d4c93c775", "messageHeadline": "added miopenGetConvolutionSpatialDim if ROCm5.5 (#14772)", "author": {"name": "Zachary Streeter", "email": "90640993+zstreet87@users.noreply.github.com", "user": {"login": "zstreet87"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0ebe8e34f801f71d6cca126e929b453fc903eb8f", "messageHeadline": "Do not use ADL to invoke std algos (#14851)", "author": {"name": "Yuriy Chernyshov", "email": "thegeorg@yandex-team.com", "user": {"login": "georgthegreat"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "70a31e047a79882ed45f22f142e537eaada8bccd", "messageHeadline": "Consume ONNX 1.13.1 in ONNX Runtime (#14812)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "2facc5efe6c537a138ddba35b7132989271a82da", "messageHeadline": "Fix function inliner bug re. outer-scope names (#14734)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 27, "deletions": 4}}, {"node": {"oid": "603026fb84df38a76ac04fd9741ef6d1fa960676", "messageHeadline": "Transpose for 16b tensors (#14877)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 175, "deletions": 2}}, {"node": {"oid": "7cd4b334a9102ff661c7d33e67a4856bc7e34daf", "messageHeadline": "[CoreML EP] Add Flatten Op and LRN Op support (#14857)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 185, "deletions": 0}}, {"node": {"oid": "bf35ad2aa34424f5d7bc72af2cdc83a1b824155e", "messageHeadline": "[Qnn EP] Call Qnn deviceCreate during backend setup and deviceFree du\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 93, "deletions": 15}}, {"node": {"oid": "d69823f76449bb6b61f8d4052be69b6a6face930", "messageHeadline": "Do not create Barrier and triggerDownstream steps if the correspondin\u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 72, "deletions": 126}}, {"node": {"oid": "c66af46fc13240bdea43bc74037977e8eba4138b", "messageHeadline": "Doc for Stable Diffusion CUDA Optimizations (#14830)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 277, "deletions": 40}}, {"node": {"oid": "c6074f3a4b09cb44b46ca4422717deb067015bfa", "messageHeadline": "OnnxRuntime QNN EP (#14791)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 7467, "deletions": 112}}, {"node": {"oid": "6044312a434987bdb2ddfd7fafbf294955f1693d", "messageHeadline": "fix TRT dockerfile documentation https://github.com/microsoft/onnxrun\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "b7fde84341f5e7e4fc8b202e9aabad4d087ec15c", "messageHeadline": "Changes to support standalone custom ops in a minimal build. (#14497)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 329, "deletions": 181}}, {"node": {"oid": "acc2ac627fc7948552aefc178e447c7860c04351", "messageHeadline": "Fp16 Activations (#14722)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 925, "deletions": 81}}, {"node": {"oid": "69c5edb11b4b8d9f5b6e522066990ac0d6e0eaee", "messageHeadline": "[wasm] upgrade emsdk from 3.1.19 to 3.1.32 (#14818)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 10, "deletions": 38}}, {"node": {"oid": "95c777b74562d3e891a329c9a388b3f5e0daa984", "messageHeadline": "Fix link to High Level Design (#11786)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "6320decf048914009a42528914986eabedb1e900", "messageHeadline": "increase Test GPU Job's timeout to 8 hours (#14850)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "79aa0acdd08e7ab84a95ce9f6d985565e68960ac", "messageHeadline": "SCELoss(SCELossGrad) support half(float) input float(half) output (#1\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1166, "deletions": 612}}, {"node": {"oid": "08699c8052124506059a5590b40bda7293ae0a11", "messageHeadline": "Address SDL warnings in recent STFT changes (#14847)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "9bdd42115c786b67dd23a99bab3ddee7c756c6ed", "messageHeadline": "add build flag for rocblas tune and fix bug (#14797)", "author": {"name": "kailums", "email": "109063327+kailums@users.noreply.github.com", "user": {"login": "kailums"}}, "additions": 15, "deletions": 4}}, {"node": {"oid": "2d079c6333d01bf0c4b67bb47cebb5683c793b10", "messageHeadline": "[js/web] disable multi-thread test on Node.js in E2E test (#14844)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "0be20dc0f60395c569eeaeaf9762521a3d1dd5f0", "messageHeadline": "Run GPU test job after all CPU test jobs succeed. (#14833)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 60, "deletions": 44}}, {"node": {"oid": "26abaeb284ed0f6f3e24d8a2bb6b74632aedeaa1", "messageHeadline": "Fix half precision gemm test accumulation error (#14842)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 25, "deletions": 9}}, {"node": {"oid": "e097e4e93c599d2a81346d6da5979ea5068d78e6", "messageHeadline": "Objective-C lib: Added support for int64 and uint64. (#14405)", "author": {"name": "Albert Grigoryan", "email": "52666717+albert-grigoryan@users.noreply.github.com", "user": {"login": "albert-grigoryan"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "dcc8fe656b131d56e208caca1fbb51dd26c9a754", "messageHeadline": "Fix type mismatch in CUDA Trilu op. (#12863)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9f6d452ca62d8fb1473129716b177b59c5c604bd", "messageHeadline": "Fix `ValueError` when testing PyTorch performance (#14450)", "author": {"name": "Ivan Komarov", "email": "Ivan.Komarov@dfyz.info", "user": {"login": "dfyz"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6603be26c6a7758524d4352e8412f9d087826bbf", "messageHeadline": "Fix prefast warnings in CUDA ops (#14814)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "3bcdb0a83abc0d643752f05bf2bc19ee005df02d", "messageHeadline": "Fix TunableOp signature generation (#14709)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 49, "deletions": 11}}, {"node": {"oid": "bc5d0c83d173acd2f897a39d916cfd8fb4a2f399", "messageHeadline": "Fix prefast scan complaints  (#14823)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "6b83ad96597c5ee2ea434ec20d8dd40844ce1f3f", "messageHeadline": "[js/web] allow unittest (onnxruntime_test_all) to run in browser (#14\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2933, "deletions": 49}}, {"node": {"oid": "a631ed77c012dbf0a86f3147e6fbe1c236f08c02", "messageHeadline": "[js/web] support flag 'optimizedModelFilePath' in session options (#1\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 109, "deletions": 11}}, {"node": {"oid": "0700788b6e38ea7aa2838d1fcab12006cb432d4c", "messageHeadline": "Disable e2e android react native CI test temporarily (#14803)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 30, "deletions": 30}}, {"node": {"oid": "702a61c3bb1fc8e8d740f7f7b487a393306153c5", "messageHeadline": "Add verbose and optimization args for parity tests (Gelu, Layernorm, \u2026", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 119, "deletions": 38}}, {"node": {"oid": "d3785ef8f623d06ee110a7457b925156bd551be6", "messageHeadline": "Fix decoder_attention scratch buffer size and prefast warnings (#14808)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "928289c41444c3a99de0faa32639fc18d0e24e2f", "messageHeadline": "STFT for DML EP (#14736)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 1484, "deletions": 5243}}, {"node": {"oid": "687118a1595465eb5a194b6562baf019baf3b146", "messageHeadline": "[ROCm] Fix Skiplayernorm error and open ke test with optional output \u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 9, "deletions": 17}}, {"node": {"oid": "abb51ec97599adee62c6c90f950a719ee778b974", "messageHeadline": "[DML EP] Fix GetInputTensor crash when accessing null tensor (#14809)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 4, "deletions": 6}}, {"node": {"oid": "29428cd9dc9b1a7a3df31ffe6a4828d551db2434", "messageHeadline": "Cjian/pr into main for 1.14.1 fix (#14805)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "d034b432ece3233b53798c8d34e654c610a3d3a3", "messageHeadline": "Add support for handling sbyte (Int8) data in C# inference tests (#14\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 9, "deletions": 2}}, {"node": {"oid": "a012d60777a26515efece4bcd228c3863733db5d", "messageHeadline": "Make MemcpyToHost to a separate stream for performance gain  (#14487)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 49, "deletions": 54}}, {"node": {"oid": "664e29627013af5ef310173b113eb1e8b6af0357", "messageHeadline": "Re-add api:javascript and api:java to the labeler (#14238)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "d925055a3ee92b4bf7df9b92973719cfd707dbb1", "messageHeadline": "Fix broken and outdated links in documentation (#14092)", "author": {"name": "James Yuzawa", "email": "jtyuzawa@gmail.com", "user": {"login": "yuzawa-san"}}, "additions": 19, "deletions": 19}}, {"node": {"oid": "16b39e5b87b7894adb6e0b936e0593ff9e3e02b1", "messageHeadline": "`symbolic_shape_infer.py`: Fix slicing a tensor that has a sympy.Min(\u2026", "author": {"name": "Ivan Komarov", "email": "Ivan.Komarov@dfyz.info", "user": {"login": "dfyz"}}, "additions": 71, "deletions": 2}}, {"node": {"oid": "b392d5351fe98c87f53b5ab1d9552f706839cea4", "messageHeadline": "Fix memory profiler (#14695)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 145, "deletions": 106}}, {"node": {"oid": "62ee0c8110882c6153921c87e36774273fe0a75e", "messageHeadline": "Migrating ORT Extensions from Git submodule to cmake FetchContent (#1\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 176, "deletions": 62}}, {"node": {"oid": "58da3cacdf16392281dad1e437ec8d1c45ccb709", "messageHeadline": "support NeoX-style rotary embedding (#14785)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1872, "deletions": 53}}, {"node": {"oid": "cad7ef93e6b19ed2c177d75ea22ee44c686ef7ac", "messageHeadline": "use python 3.9.7 in windowai packaging pipeline (#14766)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 17, "deletions": 17}}, {"node": {"oid": "5c8325a6cb5d6e9dfc1f8625503fce40d9fc2042", "messageHeadline": "Add Environment Variables for cuda tensor dumper (#14780)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 124, "deletions": 65}}, {"node": {"oid": "460b3ff4fd1633b0d83cb7ecdcbbce478c96fee1", "messageHeadline": "Update pattern matching for EmbedLayerNormalization fusion (#14344)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 439, "deletions": 21}}, {"node": {"oid": "b3b9be19b140ba073f0b1c44bfc3ffc011768cbc", "messageHeadline": "Update clang-tidy path for updated Mac image. (#14760)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "df45b12fdf8fbeb68c314b2475eabbc7c2ea3af8", "messageHeadline": "Reduce memory usage for TreeEnsemble operators (#14670)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 220, "deletions": 154}}, {"node": {"oid": "262e46e8ceac93c18439d168e73666c400da2fb7", "messageHeadline": "Update stable diffusion benchmark script (#14759)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 250, "deletions": 27}}, {"node": {"oid": "1b7f65437efa74c8cd40a85250cebb3a6aa44aa4", "messageHeadline": "Enable Opset11 Sequence Ops on DirectML, and make the CPU implementat\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1953, "deletions": 538}}, {"node": {"oid": "e9ec4c098b07a0f9d7b00888a7c37d59d02b24fe", "messageHeadline": "[CUDA] Fix FP16 Precision for Sigmoid Op (#14727)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 45, "deletions": 1}}, {"node": {"oid": "9fbb2b47424f8881e3256a867774d9325ae160c3", "messageHeadline": "Fix broken link in onnxruntime_c_api.h (#14748)", "author": {"name": "Christian Veenhuis", "email": "124370897+ChVeen@users.noreply.github.com", "user": {"login": "ChVeen"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b234df3dd060f958a7f5d329b1c75f91c6086656", "messageHeadline": "Refactor the cost check used by the transpose optimizer  (#14690)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 197, "deletions": 88}}, {"node": {"oid": "755161100a56454463a02c83ce114f2f736fe78b", "messageHeadline": "Fix CoreML API usage memory leak. (#14738)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 17, "deletions": 9}}, {"node": {"oid": "973aaf110b1d5e1501e2801172a2f1d88baf3c06", "messageHeadline": "Improve compatibility with certain STL's", "author": {"name": "Yuriy Chernyshov", "email": "thegeorg@yandex-team.com", "user": {"login": "georgthegreat"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "e6a8e6c438a3ec25c511509a9686fb623fb1bfcb", "messageHeadline": "Drop the test folder accidentally added (#14718)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 0, "deletions": 38}}, {"node": {"oid": "f76ff8c5586b305c3f19ebb47d2d3dd315f9b4d9", "messageHeadline": "Initialize bias_weight in fusion_skiplayernorm.py (#14751)", "author": {"name": "fxmarty", "email": "9808326+fxmarty@users.noreply.github.com", "user": {"login": "fxmarty"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "c0d2472edeedabbb4332432f6cee8bae61e79f48", "messageHeadline": "Disable fused causal attention (#14732)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 55, "deletions": 40}}, {"node": {"oid": "25e10f413ed99b8f1e2b7d374a60ac0b04e77d66", "messageHeadline": "[FIX] USE_COMPOSABLE_KERNEL is not defined  on ROCm5.2.3 (#14750)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "3d79b1f06eff2c8c43ab6f549a49771307fb6e56", "messageHeadline": "Create new stream for data copy for IOBidning input scenario (#14719)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 58, "deletions": 43}}, {"node": {"oid": "1ea360148f2b749074012ebe486a9324820b190d", "messageHeadline": "restore opset18 test (#14677)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 8, "deletions": 39}}, {"node": {"oid": "fbf5d09a0cecf40522ccd8b74b2f08564acb7159", "messageHeadline": "Fix random failure of ortmodule_api.py::test_unused_parameters (#14729)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ad78579b6682a98e9af3bb04ac46ee8893ed201b", "messageHeadline": "Update java/build.gradle to not use deprecated features that were rem\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 23, "deletions": 38}}, {"node": {"oid": "8372c86e7f15ed0d8d47e1e0afc115c30db39252", "messageHeadline": "[oneDNN] Update to oneDNN v3.0 (#14267)", "author": {"name": "Erick Mu\u00f1oz", "email": "erick.munoz.alvarado@intel.com", "user": {"login": "eralmual"}}, "additions": 410, "deletions": 317}}, {"node": {"oid": "7b31bcda2e9cd45c709e3fe31a544297db37ea3c", "messageHeadline": "Disable LazyTensor-ORT Test (#14703)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 0, "deletions": 21}}, {"node": {"oid": "0dda42b46cd491f5b658d291078cd81388c1db5a", "messageHeadline": "Enable some ops for QDQ Node Unit support (#14701)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "892f59b31a717a975aa305b0a42cddf77a547093", "messageHeadline": "Add string support to tile op (#14686)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 63, "deletions": 10}}, {"node": {"oid": "ae205a7924c6a2b61157a01adc8fd6236c8c96f6", "messageHeadline": "QAT POC tutorial (#14577)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 417, "deletions": 0}}, {"node": {"oid": "4e686a9a7d84ee05fe3db257d68c2b38ecf4828b", "messageHeadline": "Support building a QAT onnx model using onnxblock (#14551)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 22, "deletions": 3}}, {"node": {"oid": "ff3aed85404ba76debec38d4e16e6376dcb53cbe", "messageHeadline": "fix error due to () not used on operator priority. (#14699)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "6f99fb9d4be8921c5dc06a184b4f43dddfba5e1b", "messageHeadline": "Stable Diffusion CUDA Optimizations Part 5 (#14706)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 85, "deletions": 2}}, {"node": {"oid": "0f9d2432d21562ad93a53e8eeda048130add0bac", "messageHeadline": "[ROCm] Add WarpWise Softmax into SoftmaxTunableOp (#14612)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 234, "deletions": 102}}, {"node": {"oid": "601aab2ac8402259ac1857eb734562f48258971e", "messageHeadline": "[Testing] Arrange parity utilities for onnxruntime parity tests to se\u2026", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "68db1b62a8d235633fff4f55971c727bd26da7f9", "messageHeadline": "add noexcept to `InitApi()` and `GetApi()` (#13869)", "author": {"name": "Dale Phurrough", "email": "dale@hidale.com", "user": {"login": "diablodale"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "eb2ac72fa90807f95f3f30dfba8e6c2d2adefb80", "messageHeadline": "Stable Diffusion CUDA Optimizations Part 4 (#14680)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1121, "deletions": 211}}, {"node": {"oid": "5605c3d4540099dc701a944db42079486970a987", "messageHeadline": "Make some variables constexpr in orttraining/orttraining/training_ops\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "733ca85b7395a83c6b671936db94be98d600e3ae", "messageHeadline": "Cfu fp16 (#14538)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 3091, "deletions": 27}}, {"node": {"oid": "b1abb8c656c597bf221bd85682ae3d9e350d9aba", "messageHeadline": "skip col2im_pads test (#14685)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d49cea05fa4ae92799a057f2b7906fab09932beb", "messageHeadline": "[ROCm] Support for gpt2-based model inferencing (#14675)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 175, "deletions": 95}}, {"node": {"oid": "a216c9a3fa73ddd593f024b737dbf4aff0b06d68", "messageHeadline": "Offline tuning (#14558)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 868, "deletions": 9}}, {"node": {"oid": "f638c5a2ae3682eb3b72a2180f63ce2095450ef5", "messageHeadline": "Stable Diffusion CUDA Optimizations Part 3 (#14646)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1355, "deletions": 253}}, {"node": {"oid": "6eeeecf07fd67e410c94bfdbe43a98d045d512fd", "messageHeadline": "Remove erroneous function cast (#14673)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "2a4c9a5cbf67551ed8fd0bd4a5161a3b1861191a", "messageHeadline": "[T5 optimization] fuse rel_pos_bias and remove extended mask (#14645)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 162, "deletions": 11}}, {"node": {"oid": "37033975bbc8fee7f32073b217654f308c529ccd", "messageHeadline": "[MIGraphX EP] Add support for Mod OP (#14647)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7e9f543af4f8c3968ad17525af8b01b2bc75efd3", "messageHeadline": "Use miopenGetConvolutionSpatialDim if ROCm5.5 (#14483)", "author": {"name": "Zachary Streeter", "email": "90640993+zstreet87@users.noreply.github.com", "user": {"login": "zstreet87"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "ce9a71620f1790064521c659575851e63ddac15a", "messageHeadline": "Fix DML release build (#14661)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 18, "deletions": 13}}, {"node": {"oid": "6e1008877e45ad63bfe519e770fadfc1c1624cd2", "messageHeadline": "Update OrtEnv class documentation (#14650)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "50fa15129877355c98941fcb14eb2b15e668268e", "messageHeadline": "remove device_id parameter out of ExecutionProvider::GetAllocator() (\u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 321, "deletions": 349}}, {"node": {"oid": "22de2798f2e65fab984ea1cdf175e4c6aa40f270", "messageHeadline": "Update typing hints to support python 3.8 for training apis (#14649)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9, "deletions": 4}}, {"node": {"oid": "326cf2f5e96ce20dda90b3e3de70b280635644ed", "messageHeadline": "[ROCm] add Softmax Tunable Op (#14541)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 625, "deletions": 73}}, {"node": {"oid": "12d91173c4478e0975771c06fd9d062a33c46339", "messageHeadline": "Add extra include to fix build w/ CUDA 12 (#14659)", "author": {"name": "Misha Chornyi", "email": "99709299+mc-nv@users.noreply.github.com", "user": {"login": "mc-nv"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "ba00f3a134b054e999747a2fddd7f46505462a00", "messageHeadline": "fix problem of reduplicate input names (#14163)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 66, "deletions": 12}}, {"node": {"oid": "0de4bc70500bbf9253bbfc3f7e7bf621f3a8bd3f", "messageHeadline": "add symmetric quant in softmax (#14640)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 38, "deletions": 8}}, {"node": {"oid": "9bd022b8be1f6835a87b90297fa3b13144d53430", "messageHeadline": "Add TuningContext for TunableOp (#14557)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 760, "deletions": 294}}, {"node": {"oid": "9a9d45fefbcf2d82ca6417eaa734ba558ac6a0ee", "messageHeadline": "Add instructions for previewing docs changes (#12528)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "94bc0fe0292edab1ee26177eaf45ebb1dfefa22b", "messageHeadline": "Skip all training opset model tests (#14636)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 30}}, {"node": {"oid": "02bba3e2681ac73b0c03d3ea01ebc7b9c0eb62eb", "messageHeadline": "Switch to a static local variable to avoid global constexpr warning (\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 12, "deletions": 7}}, {"node": {"oid": "23f0e442659c34010bd40aa59490b4e7933a8b3d", "messageHeadline": "Fix SAL annotation in private DML EP interface (#14639)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c5b485d25f3dca9e512d47afc010b17d8565a6d2", "messageHeadline": "[prefast:Warning]: C26451 (#14628)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "b53038b6a097a5b7168a8783c9659cedf85c6e00", "messageHeadline": "Fix softmax block forward with small element size (#14475)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 83, "deletions": 88}}, {"node": {"oid": "875a7791bf610f9d5cc670ad20dfb58627406aba", "messageHeadline": "[DORT] Update import path (#14605)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "96b95a24eeec2c760dfd08b7ac98868860a890c0", "messageHeadline": "Add rust bindings (#12606)", "author": {"name": "Boyd Johnson", "email": "boydjohnson@users.noreply.github.com", "user": {"login": "boydjohnson"}}, "additions": 6076, "deletions": 0}}, {"node": {"oid": "30ec8b038f69cdafa59ec851258be31b8c4448c3", "messageHeadline": "Test and fix optimizers LayerNormFusion, BiasSoftmaxFusion, Transpose\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 605, "deletions": 228}}, {"node": {"oid": "cfda876a3f0f79cc88e57996034acca19d81d30b", "messageHeadline": "Remove torch package from requirements.txt of stable diffusion models\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "0a6b22018fd52d927c745cfa1d51d9c7db629f69", "messageHeadline": "Move TRT include_directories to outside scope (#14622)", "author": {"name": "Kevin Chen", "email": "45886021+kevinch-nv@users.noreply.github.com", "user": {"login": "kevinch-nv"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "767619cf3b823f7ae803cd5307efd4fb7cfa9546", "messageHeadline": "Rework C API to remove new/delete warnings (#14572)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 54, "deletions": 46}}, {"node": {"oid": "10ab2529820e0e5ee179180855d253769656b447", "messageHeadline": "Enable parallel output reordering in MlasReorderOutputNchw() (#13643)", "author": {"name": "Alex Kogan", "email": "82225080+sakogan@users.noreply.github.com", "user": {"login": "sakogan"}}, "additions": 164, "deletions": 60}}, {"node": {"oid": "ba8a00f62f96ce93eeae2216353a33832b610742", "messageHeadline": "[TVM EP] Support zero copying TVM EP output tensor to ONNX Runtime ou\u2026", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 123, "deletions": 47}}, {"node": {"oid": "0b52a887b6a20d6a91e0d62128206eb3844d5077", "messageHeadline": "[Readme] Update table for build pipelines (#14618)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 9, "deletions": 8}}, {"node": {"oid": "8ed3dfe0633f5772f1684a5c9f7e68c7acec00d2", "messageHeadline": "Revert \"try VS 2022 in windowsAI pipeline (#14608)\" (#14619)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "e9ab56fa64c0644a2dc5287d1cd3b945bf7d7981", "messageHeadline": "Adding RunOptions synchronization behaviour to C/C++ API (#14088)", "author": {"name": "Maximilian M\u00fcller", "email": "44298237+gedoensmax@users.noreply.github.com", "user": {"login": "gedoensmax"}}, "additions": 12, "deletions": 9}}, {"node": {"oid": "cd7098fdf47e804ff84ca0cc2522e82bce8a2e7f", "messageHeadline": "fix snpe build (#14616)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8f34c8c8eda2890b502021283c463d0552e45b8d", "messageHeadline": "Introduce collective ops to ort inference build (#14399)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 764, "deletions": 277}}, {"node": {"oid": "b539c364eec10cf238ffaf2e323ababaa05e433b", "messageHeadline": "Some kernel changes for TULR  (#14517)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 802, "deletions": 132}}, {"node": {"oid": "b6bec543415f16eaf8b687d016e425f5c470cd48", "messageHeadline": "Revert mimalloc from v2.0.9 to v2.0.3 (#14603)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 4, "deletions": 43}}, {"node": {"oid": "585f43e31ddd70ce0cae77cd1207e2752999a12e", "messageHeadline": "Remove Identical Children Consolidation from default transformer uiti\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "8de885fdb1b070dc85e19c92f1dccdeaee482131", "messageHeadline": "reduce cuda library binary size (#14555)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 61, "deletions": 129}}, {"node": {"oid": "742658d171fdabb9032f4f2e5a847687188231b6", "messageHeadline": "Stable Diffusion CUDA optimizations Part 2 (#14597)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 386, "deletions": 61}}, {"node": {"oid": "f88a4646cdb47f106fc0838426dabe362353321d", "messageHeadline": "try VS 2022 in windowsAI pipeline (#14608)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "cf8bad7f19830567bde7725835501c0d0dda7a71", "messageHeadline": "Fix CI failure: temporarily disable real model tests from onnx repo (\u2026", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 12, "deletions": 1}}, {"node": {"oid": "d632f9a3fa126a64bcab7be2000220e2a73765be", "messageHeadline": "[ROCm] Enable Sampling Op UT on AMD (#14581)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 137, "deletions": 113}}, {"node": {"oid": "a5dab850b8dfa64d8818eefa01e0df40cf3e4400", "messageHeadline": "Bump jszip from 3.7.1 to 3.8.0 in /js/web (#14536)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "6f2dd10d5278a85f85e552b37b3b469ab198d31b", "messageHeadline": "IdentityBuilder should add Delimit for each input (#14592)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "b8fb9320ac55c32fa8b5bfb1a4a0335537fac048", "messageHeadline": "[DML EP] Fix ScatterElements registration (#14560)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "20684021dad9339941e62618ec241ddc6ecc0c43", "messageHeadline": "do not use raw pointer for CpuBuffersInfo::buffers (#14574)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 4, "deletions": 6}}, {"node": {"oid": "4bb95d7690bac6b25622dbef5b711c15ffb00eee", "messageHeadline": "Change the return type of softmax function to Status (#14559)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 69, "deletions": 71}}, {"node": {"oid": "3d7518762ace6929be98e1203174c2dbf1ac094e", "messageHeadline": "[ORTModule] ATen Support for upsample_bilinear (#14519)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 58, "deletions": 9}}, {"node": {"oid": "c1a0fc55e7a5d685917a966f4059970869962986", "messageHeadline": "[ROCm][MIGraphX EP]Add back in support for gfx1030 (#14565)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "999e5bf45ee9ccdfc6b30b93468e8c9d7ea0bc7f", "messageHeadline": "Add SLN support for t5 model with beam search (#14429)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 130, "deletions": 10}}, {"node": {"oid": "638f21b969143f65f4c47ce6b0f41f435e1d3ec3", "messageHeadline": "Upgrade doxygen to fix C API docs build issue (#13950)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 14, "deletions": 58}}, {"node": {"oid": "a5eb616819584d8f15ca21124e3892e9f209abc4", "messageHeadline": "Enable ability to control whether or not to quantize the bias (#14549)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "7eca42484ce2afc04cedd50658abed1c0485a8dc", "messageHeadline": "link mpi when either use_mpi or use_nccl enabled (#14467)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 22, "deletions": 15}}, {"node": {"oid": "c6c11039d76c95798238e358c898a562a200a239", "messageHeadline": "Fix sharing scalar bug (#14544)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 98, "deletions": 7}}, {"node": {"oid": "a6c5ba01851ace73491a3f8edcb5743c747dc78d", "messageHeadline": "Stable Diffusion CUDA Optimizations (#14428)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3263, "deletions": 328}}, {"node": {"oid": "1059cf6d986787716081404f6b0a15c309becf46", "messageHeadline": "[ROCm] Fix ROCm build issue caused by REMOVE_ITEM  incorrect path (#1\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 4, "deletions": 6}}, {"node": {"oid": "7b75ebdb31ddff0cb7c210d1909be7e116b0c077", "messageHeadline": "Bump http-cache-semantics from 4.1.0 to 4.1.1 in /js/web (#14535)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "78280bf565082141693008d8ca58805697fe45f4", "messageHeadline": "fix build err inbuild with minimal_build conjuncting disable_exceptio\u2026", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 32, "deletions": 0}}, {"node": {"oid": "20b5a75cfa552a6280dd391b83ec81f48998e36e", "messageHeadline": "Add missing semicolon (#14143)", "author": {"name": "Pavel Grunt", "email": "xerus@users.noreply.github.com", "user": {"login": "xerus"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "549cbc7e69433433cd464efcb6bc78e796fbae3b", "messageHeadline": "Fix issue with schema lookup where there are custom ops using the ONN\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 115, "deletions": 86}}, {"node": {"oid": "cfb6e528c84b1d9e7945cbe495d8c9f4c52afcef", "messageHeadline": "[js/web] remove 'module' field from package.json (#14532)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "03cfb7d73ede48345d90688071232bd76e813e1f", "messageHeadline": "Use ORT node names in DML graphs/ops (#14461)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 131, "deletions": 2}}, {"node": {"oid": "0bcca7ad4537a2c6ebf5a7b5a01f9077fa0992d4", "messageHeadline": "Fix Gather to Split optimizer (#14478)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 170, "deletions": 28}}, {"node": {"oid": "3d8fa4d77bc621109ac0e33014195e0b6dbd621b", "messageHeadline": "GetTrainingApi to not print to stderr when not an ort training build \u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 23, "deletions": 15}}, {"node": {"oid": "68a402e739abd596e6fa8d05a3e185626aa1b669", "messageHeadline": "Add support for python 3.10 for onnxruntime-training cuda and cpu (#1\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 19, "deletions": 1}}, {"node": {"oid": "01cafe89f07ba0c4492b5042ab2dfe98b52c3960", "messageHeadline": "Specify deps in deps.txt and manifest (#14530)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 678, "deletions": 12}}, {"node": {"oid": "62442c3d2747c2291c8cf9367dcab803e223dda2", "messageHeadline": "Enable multiple step run for adamw tests (on device training)  (#14520)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 20, "deletions": 39}}, {"node": {"oid": "7954976e0a291b60872e8e144a9e94d9b26aa3cd", "messageHeadline": "Fix python packaging pipeline (#14533)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0578eeff9196b7ab9bd6deaae8e8d92b3d9ee6ee", "messageHeadline": "upgrade EsrpCodeSigning from v1 to v2 (#14531)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "61e7636e618514c1d59c6d46bf7ccaa3df7fda88", "messageHeadline": "Re-work GetAvailableProviders API (#14486)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 97, "deletions": 53}}, {"node": {"oid": "d9e675a2af4cf1d528f98ba0c98b54a6d251729b", "messageHeadline": "Do not fuse DQ+Node+Q if DQ produces graph output (#14509)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 64, "deletions": 0}}, {"node": {"oid": "3d388a1aeada7d9b7b334ba915d684d3ac3c2066", "messageHeadline": "change deepspeed version in warning from 0.7.3 to 0.8.0 (#14527)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b1a17188a69be90e4f8e87c89dce631377085f06", "messageHeadline": "[js/web] add LRN unpacked kernel for webgl backend (#14459)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 108, "deletions": 3}}, {"node": {"oid": "bc8686d3659566f3fe2942c7afc66e654383292e", "messageHeadline": "Fix prefast warning 11263 (#14508)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "3cc81460e045daac6edf22175ca143fb3b5047e6", "messageHeadline": "Register ScatterElements-16 (#14425)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "a1ed703b73430eee1f76c2957e6f6a485b451a92", "messageHeadline": "Fix Python format CI failures due to black && pyright package updates\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "6fa4555a06e8b5ec588ff1d7880ff42eb6ba2007", "messageHeadline": "Including support for Deepspeed 0.8.0 (#14506)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "d06ad9462b680163e4d83f8a294979ba23114bd6", "messageHeadline": "[Bug Fix] Include python training apis when enable_training is enable\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 43, "deletions": 31}}, {"node": {"oid": "d1533c27eb98759f4747a3b05a6d9d0ca538a4ca", "messageHeadline": "[oneDNN] Improved thread handling  (#13618)", "author": {"name": "Erick Mu\u00f1oz", "email": "erick.munoz.alvarado@intel.com", "user": {"login": "eralmual"}}, "additions": 692, "deletions": 94}}, {"node": {"oid": "596b505f6a3a27359786a8513576e5efbb0cf1df", "messageHeadline": "Add spin args for perf test (#14507)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 23, "deletions": 3}}, {"node": {"oid": "80f807c03d17fe398b557bc770b554dfaaa44fd6", "messageHeadline": "upgrade protobuf to 3.20.2 and onnx to 1.13 (#14279)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 157, "deletions": 26}}, {"node": {"oid": "e2dd1315c779f9f60aedd0243d61af625fdf5cc3", "messageHeadline": "Fix build for --enable_language_interop_ops + DISABLE_ABSEIL=ON (#14469)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 12, "deletions": 8}}, {"node": {"oid": "a5b620e79d39a78b630e9f915e03f411889d3393", "messageHeadline": "[Build] Fix arm64 Docker build (#14283)", "author": {"name": "Ankit", "email": "ankitmaurya001@gmail.com", "user": {"login": "ankitmaurya001"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "679ae7ff3352b31fce486122f73495a46d5ed30d", "messageHeadline": "[Java] Fix warnings (#14076)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 27, "deletions": 15}}, {"node": {"oid": "764202d7400e1c610d42a4260b3a5994f4987d89", "messageHeadline": "fix prefast warning (#14446)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3b6d551c3563738d6396e9e44715ecc3ffc4678b", "messageHeadline": "Enable ccache for HIP objects (#14465)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 34, "deletions": 0}}, {"node": {"oid": "7aecb2150f6a8b4d9c70a4553b8943b06ae43b9e", "messageHeadline": "Fix onnxruntime-CI-nightly-ort-pipeline Failure (#14464)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "91d42e9d85a0e9832bbad982462a4989a5b01d7a", "messageHeadline": "Tool to Convert ONNX Model to TFEvents (#14160)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 398, "deletions": 8}}, {"node": {"oid": "d9219685ade750211a887f9cb217580eefddef24", "messageHeadline": "always set OpSchema in CreateNodeHelper() (#14356)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 12}}, {"node": {"oid": "b5b70eaa8cdcfba18eb95dc872f9bba62335d73a", "messageHeadline": "Bump ua-parser-js from 0.7.31 to 0.7.33 in /js/web (#14435)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "f87dd408f67c460a5d0ab0280b420870d36cc50f", "messageHeadline": "Support long sequence in attention (#14371)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 195, "deletions": 3}}, {"node": {"oid": "368d2fc11ede1d503ef2d2c247852b48b3fcadd0", "messageHeadline": "Added E2E test for Image Tensor API (#14406)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 72, "deletions": 7}}, {"node": {"oid": "4ef64f3681819924b93d9b844ab847d2eb4b4fad", "messageHeadline": "Fix warning c26409 (#14079)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "de11527d76d4613c53fbbe11188e9cf525f8b4b5", "messageHeadline": "[js] fix js/web bundle (#14434)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "eacd829d230e7419594c3c06dfa82a084f50429f", "messageHeadline": "Bump ORT version number (#14226)", "author": {"name": "Rui Ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 27, "deletions": 18}}, {"node": {"oid": "d9c744ed9a137f5ff2ff046edd20221ae6f7939e", "messageHeadline": "Fix a bug in t5 beamsearch with half precision (#14436)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2b1a59f01abd38d5fd40d75b3f41547791980fbc", "messageHeadline": "cpu support of LpPool(18) (#14205)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 417, "deletions": 5}}, {"node": {"oid": "edb377f2cbc96bdd1753c09e9575226337338c7b", "messageHeadline": "[DML EP] Upgrade DML to 1.10.1 (#14433)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "3b8dfe2e279381d181e7938e4591d5964cc2649a", "messageHeadline": "Don't use free to satisfy Prefast requirements (#14354)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 22, "deletions": 28}}, {"node": {"oid": "4d9ddb51937e37841c7118fc61f38773d4124e96", "messageHeadline": "[js] upgrade packages in js/web/test/e2e (#14334)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "32c05fcdd12636ef06e2ca3875704183e3da293a", "messageHeadline": "Add Col2Im CPU op (#12311)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 317, "deletions": 1}}, {"node": {"oid": "94b17919745b3c8a86000554871a0a9a3dee5fbf", "messageHeadline": "Upgrade CUTLASS to v2.11 and add sequence length threshold for cutlas\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 37, "deletions": 13}}, {"node": {"oid": "7cc9aed314cb32d4e12b44186074e5c52e29f6b3", "messageHeadline": "Android package custom build script update (#14403)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 118, "deletions": 32}}, {"node": {"oid": "3bc092b1eaedf30be9cef5388f1a2f026ec6800c", "messageHeadline": "Update ORT format v5 change docs to cover limited backwards compatibi\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 23, "deletions": 8}}, {"node": {"oid": "85d7e9c5960f545c21246c8e9140bd1ab403949a", "messageHeadline": "Fix unused variable for CUDA EP builds with USE_FLASH_ATTENTION off (\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "3c1ef7dee6d910f3d570a28a8fb541d0325ec58d", "messageHeadline": "Fix CI build with no Abseil. (#14400)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "81120e9e8b377567daa00d55614c902f35b2ae8f", "messageHeadline": "Add custom tolerance option for onnx_test_runner (#13683)", "author": {"name": "Kevin Chen", "email": "45886021+kevinch-nv@users.noreply.github.com", "user": {"login": "kevinch-nv"}}, "additions": 33, "deletions": 3}}, {"node": {"oid": "7b6d880b28db5e52ddaecb0f09533027daab14f2", "messageHeadline": "cpu to support bitwise ops (#14197)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 256, "deletions": 2}}, {"node": {"oid": "77b455b969db15d911cd6de5009f3a30fb42c531", "messageHeadline": "Ort openvino 4.3 cli (#14341)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 46, "deletions": 118}}, {"node": {"oid": "c252a7f99208163a7a0cb8c445ca109ec7d2f903", "messageHeadline": "Remove exclusions for ONNX model tests that now pass. (#14337)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 13, "deletions": 43}}, {"node": {"oid": "05915d8393dd3bd0378731ab9ed7c47186554a4e", "messageHeadline": "support Pad(18) (#14219)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 103, "deletions": 34}}, {"node": {"oid": "f03c507cf05a1387738bc0d2f75d220511e22758", "messageHeadline": "Fix fuzz test (#14385)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 10, "deletions": 9}}, {"node": {"oid": "abaed6f474fdf166a3c56fb25c7a6207f19acbef", "messageHeadline": "Add link to Python API examples (#14345)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "a95fcb4345a929d7122bbb18b2e54e353031cc85", "messageHeadline": "UNet fusion and fp16 conversion for stable diffusion (#14248)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 632, "deletions": 45}}, {"node": {"oid": "e57c312f9d3c61509bf8e41d9f13e53224fe1a7f", "messageHeadline": "Pin sphinx to avoid broken link (#14383)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "cf3661ff6d19157d795b0b0bd58a3ea04174ca36", "messageHeadline": "Revert \"Allow PostAnalysis@2 task to continue on error for Windows_Pa\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "0d401196246152b625c2c0d5069273add80c9e93", "messageHeadline": "Fix broken link (#14368)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "de7a868d5f3390d7c095a53c26abd39f402f3f93", "messageHeadline": "Update quantization_defs.cc (#14380)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "2d8ee5251cd7037d3d94cd22ada77e10ba95f192", "messageHeadline": "Misc transformer fixes - 3 (#14320)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 40, "deletions": 12}}, {"node": {"oid": "72821a61132c282c058e643e227cbf19504d0048", "messageHeadline": "Add PyTorch 2.0 to ORT transformer benchmarking (#14300)", "author": {"name": "kunal-vaishnavi", "email": "115581922+kunal-vaishnavi@users.noreply.github.com", "user": {"login": "kunal-vaishnavi"}}, "additions": 42, "deletions": 5}}, {"node": {"oid": "414b012f42e3f9a343f827c363b30b94011dc3f7", "messageHeadline": "Add memory efficient attention from CUTLASS (#14343)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2140, "deletions": 75}}, {"node": {"oid": "e64f357ad484e790450e0c00f1eae0a9d5f77e35", "messageHeadline": "Fix some prefast checking found problems. (#14342)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3b382ea7e17e24f17a8bfb0522b097be2dd03bc0", "messageHeadline": "Free OrtStatus in ASSERT_ORT_STATUS_OK, make run_android_emulator.py \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 47, "deletions": 52}}, {"node": {"oid": "22fdc31667cd0219a7cb69043a777415acb270d3", "messageHeadline": "remove unnecessary waitOnEPStep when current node and the consumer no\u2026", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 248, "deletions": 86}}, {"node": {"oid": "cd24f0794a22523298d9a5cab489a7a2f6ccc128", "messageHeadline": "Extend ort_backend.py for another ep (#14349)", "author": {"name": "Kyushick Lee", "email": "kyushick@gmail.com", "user": {"login": "kyule7"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "3d6cea14f4e61f6dd48a4fd4e2dc3acb424cbbe9", "messageHeadline": "Remove intermedia obj files once build finished (#14361)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 15, "deletions": 1}}, {"node": {"oid": "668586e8f800411ad6088b7a5ec11563b25bacac", "messageHeadline": "Support muP in Attention (#14348)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 110, "deletions": 32}}, {"node": {"oid": "1dd07d147da4e2011873a9beda17faa78e5ab926", "messageHeadline": "fix windows build error (#14362)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "432a9912a35ad9d34e12e73407952edc02c047a1", "messageHeadline": "Fix LORT CI failure due to PyTorch change (#14367)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 34, "deletions": 1}}, {"node": {"oid": "36ba3d8d2176df3c17cbfe875f01cdedbfa1343e", "messageHeadline": "Exclude a multi-stream case from reduced ops build (#14351)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "5d6a04914112ab6347b62dc1af57fb9d05381f2e", "messageHeadline": "support ScatterND(18) and ScatterElement(18) (#14224)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 218, "deletions": 15}}, {"node": {"oid": "d2c3d8eb38fc0715c36fcf851030d038edc7ba58", "messageHeadline": "Add Bert/GPT2 fusion change for new attribute mask_filter_value in OR\u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 54, "deletions": 13}}, {"node": {"oid": "ae0e090c7b93dae4e71e3fa030e4757b18786da3", "messageHeadline": "Fix post merge jobs pipeline build issues (#14346)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 205, "deletions": 180}}, {"node": {"oid": "ea7bbd667d14332a9c8f1c4f6e832a1663296773", "messageHeadline": "fix headers for training apis (#14350)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 33, "deletions": 32}}, {"node": {"oid": "b51415b0ea24f404417fc0d37c3c7bcf48ae5042", "messageHeadline": "disable cache for training_x64_debug (#14358)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "80d61989e9339770cba5c1ce6cf5f742cac4cd25", "messageHeadline": "Unit test modification for TensorRT EP (#14339)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 3, "deletions": 7}}, {"node": {"oid": "a491f33f54b92264ac5059d1c0dd133d23c85dcd", "messageHeadline": "Allow PostAnalysis@2 task to continue on error for Windows_Packaging_\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "20e164786e8e61e630d8d5f7024cbd484d77d1ff", "messageHeadline": "[objc] Fix parameter name in documentation. (#14330)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "904e63633a5ba3c1b7e5ff1fa0030bb8307e2f2e", "messageHeadline": "increase the time limit as more unit tests added (#14327)", "author": {"name": "Rui Ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c4e693c4b7920d8107499858f725d64713ac4ef7", "messageHeadline": "update gsl-lite license (#14318)", "author": {"name": "Rui Ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "3c695f78fea656d1d56876282b75ed3fca8b979a", "messageHeadline": "Bump electron from 15.5.5 to 18.3.7 in /js/web (#13617)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 16, "deletions": 16}}, {"node": {"oid": "de17d53c504c7edea56d102a703a395308d6a2a5", "messageHeadline": "Custom Op runtime wrapper (#13427)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1500, "deletions": 37}}, {"node": {"oid": "734ae398eeebfff2f2c4996d660725d63dffe59e", "messageHeadline": "Fix a security warning in cuda gemm int8 kernel (#14335)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "8236808e89b122a7126eb593e6073d33cfacc00a", "messageHeadline": "add opset18 node test (#14236)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "dab900dfa0f90b74c7580c9cb31b67064ca77b16", "messageHeadline": "Fix type mismatch when ORT_ENABLE_STREAM is off (#14324)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "f6d0598b4d92a4666d14cf3a3cc6d71790110047", "messageHeadline": "DML EP return clearer error message when users attempt to use softwar\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "477cad30514f5f239aa6d7a6640e43d79f65b199", "messageHeadline": "[CUDA] Add trt cross attention kernels (#14328)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 233399, "deletions": 1841}}, {"node": {"oid": "a8df6c35f84cd8136000df1d80e504e055ce2363", "messageHeadline": "Support flash attention on 2d attention mask for gpt2 left padding. (\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 61, "deletions": 3}}, {"node": {"oid": "30b9f5dde14aefb6c5761bebe0cb2884e10fe913", "messageHeadline": "Clean up TensorRT deprecations, warnings, unbounded string copy (#14148)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 71, "deletions": 70}}, {"node": {"oid": "60290393f3d3afe240b5a5886a5da2b8e2dcb835", "messageHeadline": "enable ort-extensions in wasm release builds (#14239)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 16, "deletions": 2}}, {"node": {"oid": "2db57a53a304f294844dbdea15bdeae0dd628917", "messageHeadline": "Add mask_filter in Attention related ops' attribute (#14274)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 161, "deletions": 66}}, {"node": {"oid": "caa5900508005e2405b5c3c370aafdf1fbf4e5b4", "messageHeadline": "delete unused local typedef VK to fix pipeline error (#14322)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "19b4d9d41e83d6770dc2955d77787c7e81c96a68", "messageHeadline": "Fix murmurhash3 inclusion in TensorRT shared library (#14221)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 12, "deletions": 374}}, {"node": {"oid": "c99cd06b10b53d83ccc4c4ffedc34b2950df2780", "messageHeadline": "fix transformer model unit tests (#14319)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "909d7f4be5037043c4c1eec163a90c7ded1639d6", "messageHeadline": "Skip some tests to pass orttraining-gpu with TRT8.5 (#14314)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "db690793126a76d3aa5428238b5d6b54b999e2e6", "messageHeadline": "fix roctracer missing nccl operation bug (#14277)", "author": {"name": "kailums", "email": "109063327+kailums@users.noreply.github.com", "user": {"login": "kailums"}}, "additions": 17, "deletions": 0}}, {"node": {"oid": "d95249f516e83a7f4465604e8ef1c4f2ed02635c", "messageHeadline": "Removing Double QDQ from Graphs (#14024)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 394, "deletions": 9}}, {"node": {"oid": "fb801d58b196d21b7e35012823ae416815cbc3da", "messageHeadline": "Add Cache in Linux CPU Aten Pipeline (#14313)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 40, "deletions": 10}}, {"node": {"oid": "3a9f30df468812d604b53ca1e0a01fa5d9fd103e", "messageHeadline": "Compatibility patch for nlohmann/json < 3.9.0. (#12394)", "author": {"name": "xkszltl", "email": "xkszltl@gmail.com", "user": {"login": "xkszltl"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "7f374f40125bc2f57a43d18033878c64d1d4d5e8", "messageHeadline": "Fix build error on Windows if Python debug libraries are installed (#\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 15, "deletions": 2}}, {"node": {"oid": "49cfb56cc3d405a37047de5b26b2dfdf6198a72a", "messageHeadline": "Fix subgraph index issue in TRT (#14305)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 120, "deletions": 166}}, {"node": {"oid": "6d60dc24fe811a1f4a84fc424192bfe95df85f30", "messageHeadline": "install shared deps script (#14234)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 67, "deletions": 212}}, {"node": {"oid": "fe052e603bfb1e07e8db593dc3d98138961946fd", "messageHeadline": "ROCm header path updates (#14170)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "9c0c49900ef4caca50125fbb0517322af26a3d8d", "messageHeadline": "Perform QlinearConv for a batch in a single parallel (#14296)", "author": {"name": "spampana95", "email": "sai.kishan.pampana@intel.com", "user": {"login": "spampana95"}}, "additions": 71, "deletions": 1}}, {"node": {"oid": "15141a40b4a67c02197a7cc7f93abad24b1b4f6b", "messageHeadline": "Add present_past_share_buff to QAttention Defs to enable QAttention r\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "2a82f95040d739e202d74ffff41e15ce8ab0404e", "messageHeadline": "Increase package python test pipeline timeout limit (#14288)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "bd39c8f35e5ef061ef55c615f41dcf7d11c39cdd", "messageHeadline": "Fix causual flash attention related kernel run (#14299)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8824f812e07603e36b2c9ba0d7f7593b008e5db4", "messageHeadline": "optimize topk for greedysearch (#14271)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 427, "deletions": 90}}, {"node": {"oid": "4f309f05ca580e7ee641cb6dcdd95d5de850b21b", "messageHeadline": "[CPU] Resize of Opset 18 (#13890)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1577, "deletions": 181}}, {"node": {"oid": "f0555eb4379715c2a9d5ba7bbdb224233f58f14e", "messageHeadline": "Improved test cases by using paramerters (#14246)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 15, "deletions": 2}}, {"node": {"oid": "6ac7c894bf1f7c9530fffb305ff621f2f3199c6f", "messageHeadline": "[bug fixed] use different node names for different dedicated QDQ pair\u2026", "author": {"name": "Yilun Huang", "email": "hyl1024@yeah.net", "user": {"login": "HYLcool"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "114f18357a73bf6cada92b1047b3403e5893014d", "messageHeadline": "Add Java and Objective-C bindings for RegisterCustomOpsUsingFunction.\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 152, "deletions": 17}}, {"node": {"oid": "a909cc0e1b63b865bd5a0aa194b74ac0cda412dc", "messageHeadline": "Improves parallelization by trees for TreeEnsemble (#13835)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 178, "deletions": 128}}, {"node": {"oid": "d3a09cf77f6e2ca3f55108ee8aef0d79d1b30ae0", "messageHeadline": "[ROCm] use pytest-xdist for fast pytest (#14261)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "9bd92069284d2b9f7866aa6b1af96651a018fd0e", "messageHeadline": "Attempt to fix flaky Windows GPU CI Pipeline 'cuda' stage. (#14281)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 29, "deletions": 27}}, {"node": {"oid": "16e39807e09b81b2714bade5dbe5b9180e1a2a77", "messageHeadline": "presence_mask should be sampling only (#14275)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "cc7799835e9840d61e0ff16823b20e2be9186955", "messageHeadline": "Enable a single build with optimized inference and on device training\u2026", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 67, "deletions": 41}}, {"node": {"oid": "fb3c1221e42b4a739e0223d9ff070e76562b87ab", "messageHeadline": "Fix Prefast Warning (#14250)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "ea12b674c0d76bcc3dc004a8d31ccf279a0c8a2b", "messageHeadline": "Disable the failing opset 18 model tests that are breaking the packag\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 67, "deletions": 1}}, {"node": {"oid": "c9a53c925510a101f5ca94d5ecda0924e40a8463", "messageHeadline": "Some changes to Sampling Op (#14218)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 236, "deletions": 23}}, {"node": {"oid": "3898b22a1af2fcf6705cf8444ec05fe1989a758a", "messageHeadline": "Fix some prefast warnings (#14247)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 26, "deletions": 5}}, {"node": {"oid": "dee36f8adea50f5670b307d2d9fdf9997346728e", "messageHeadline": "DML EP Register ScatterND-16 (#14240)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 22, "deletions": 1}}, {"node": {"oid": "8f7eb75c3ef659da5cc85100a31199c5c4e31808", "messageHeadline": "fix greedysearch token out of range bug (#14242)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5c16e0befb6ea98715d4522ed995afc7c31a1e58", "messageHeadline": "[web] utility functions for tensor<->image conversion in ORT web (#13\u2026", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 3006, "deletions": 1963}}, {"node": {"oid": "99a4036c80ce2b4ddb0b7df423544e48220f1f85", "messageHeadline": "[DML EP] Add FusedMatMul (#14196)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 24, "deletions": 77}}, {"node": {"oid": "712f781702222cc3f8fe156cdf7dd0f8addc4fbc", "messageHeadline": "Make CK an optional dependencies and only built with ck if ROCm >= 5.\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 74, "deletions": 15}}, {"node": {"oid": "b9ecd428c1092e8994262f3ce52414c305cebcd2", "messageHeadline": "Add ability to register custom ops by specifying a function name (#14\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2401, "deletions": 2145}}, {"node": {"oid": "f5b4b0f77d7fa625008d0d44c18fc8ef27e4c44a", "messageHeadline": "Add support for 'axes' attr of unsqueeze in opset 13 and add ut (#14071)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 138, "deletions": 4}}, {"node": {"oid": "7654cd50e864e75ee477982aa919992e636fdcaf", "messageHeadline": "Openvino ep 2022.3 v4.3 (#14210)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 666, "deletions": 1465}}, {"node": {"oid": "26a6b40554f50d8c5949c33bf41d93aaacd35ff8", "messageHeadline": "Delete add-issues-to-project (#14147)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 0, "deletions": 18}}, {"node": {"oid": "dd2df460b3bc717b753c315fb00c368658b49748", "messageHeadline": "Split(18) (#14015)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 247, "deletions": 37}}, {"node": {"oid": "012b34dc4e6f3cf588039b7aded6d3e65f53c7a8", "messageHeadline": "Add --use_multi_head_attention in transformers fusion (#14198)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 181, "deletions": 93}}, {"node": {"oid": "83ad562826dc4abf79be2863a64ca401af427d3e", "messageHeadline": "Rename CloudEP to AzureEP (#14175)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 183, "deletions": 183}}, {"node": {"oid": "3b79b8eb1dc9036d6050df4a2af078194f9d5911", "messageHeadline": "fix reshape fusion error in numpy 1.24 (#14231)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "52fc1ff21e00d9f20da6cb60c53b6fc681531e36", "messageHeadline": "[DML EP] Fix FusedMatMul crash when batch > 1 (#14225)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 20, "deletions": 29}}, {"node": {"oid": "a4180d79c59f872f5908979fc5be2f172543d8d4", "messageHeadline": "Multi-tensor SGDOptimizer (on device training) (#14083)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2521, "deletions": 182}}, {"node": {"oid": "ecd5ce0b338658741da843d73c6a837d8d204cfb", "messageHeadline": "Use json format to save and load partition config (#14169)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 310, "deletions": 254}}, {"node": {"oid": "7a9a6bcebd2fe987de5cb220efa0d377f1ab68f6", "messageHeadline": "Improve TopP sampling (#14192)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 120, "deletions": 127}}, {"node": {"oid": "d92c663f28d95eb36ade0b5b8f36e33b963a5dd8", "messageHeadline": "Create dedicated build for training api (#14136)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 197, "deletions": 172}}, {"node": {"oid": "3a39736a2cd0dc3350312aefbf20e565e1015961", "messageHeadline": "Bump json5 from 2.2.0 to 2.2.3 in /js/web (#14110)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 7, "deletions": 13}}, {"node": {"oid": "342a5bf2b756d1a1fc6fdc582cfeac15182632fe", "messageHeadline": "Improve rpb cuda kernel (#14195)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 50, "deletions": 18}}, {"node": {"oid": "d152452d4b02cff90138dc2f92b5f49ca003a47b", "messageHeadline": "Tune test case for hybrid cpu (#14204)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "90142899bddd5e849a48402f4986065624191b38", "messageHeadline": "Supporting Intel AMX instructions in quantized GEMM (#14042)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1991, "deletions": 32}}, {"node": {"oid": "a01bf8dbb1cf7a025921ad852332b81a1be4c1ca", "messageHeadline": "rename CrossAttention to MultiHeadAttention (#14201)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 86, "deletions": 86}}, {"node": {"oid": "6b8c72cfa6f00e02e03b428427076065dd0a6e73", "messageHeadline": "pin ort-ext to 81e7799c69044c745239202085eb0a98f102937b (#14044)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 269, "deletions": 12}}, {"node": {"oid": "f4ea781b81f1213c20eacdea703ea5f49fbf202b", "messageHeadline": "DML EP Register Identity-16 (#14053)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 56, "deletions": 2}}, {"node": {"oid": "05e26f302ab0f58f67b999de902c5858a633ef0e", "messageHeadline": "Hot fix for prefast failure to unblock python package pipeline (#14206)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "3d8b596cb9e5e3e9640c44789e591aaf0c91c72b", "messageHeadline": "Use a local copy of murmurhash3 in TensorRT shared library (#14207)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 366, "deletions": 15}}, {"node": {"oid": "da57c0a701b9098bf413604b800508713335eb74", "messageHeadline": "Add protected destructor to Provider structure (#14152)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "f8117b6f870e7623061d55f71574f2645984770f", "messageHeadline": "Add catch-all exception handler to API_IMPL_END (#14194)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "33367fa2dce8c98b4521c89501c06e9102e7d9d8", "messageHeadline": "[MIGraphX] update the MIGraphX version used in ORT to rocm-5.4.0 (#14\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "6463f4383bb481498f30cd52de5f3bcd320e0dc5", "messageHeadline": "make WITHCACHE as an option in MacOS workflow (#14188)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 17, "deletions": 5}}, {"node": {"oid": "7e751ac6e682625146dad8cc55500a15f0425063", "messageHeadline": "update convert_generation for Attention op change (#14191)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c151afec7131e76d31b675817a4b3ff4ee80cfca", "messageHeadline": "[DML EP] Fix unconnected node removal logic (#14193)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "906f578be89cc5f841f30aa5f0f08030df42f63b", "messageHeadline": "[DML EP] Update DML_FEATURE_LEVEL 5.0 (#14172)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1be36913cc68cc86f20b8d7bdfce10091f715e16", "messageHeadline": "to work with onnx 1.13 rc, implement ver 18 reduce and optioanl ops, \u2026", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 492, "deletions": 179}}, {"node": {"oid": "79dc39600f400405ea196b0903a5100b08bac7f2", "messageHeadline": "Replace distutils by setuptools to import build_ext (#14108)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 22, "deletions": 13}}, {"node": {"oid": "64541a587df11d3e54d54c978a5ca204b8869721", "messageHeadline": "[DML EP] Remove unconnected nodes from the graph (#14155)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 127, "deletions": 15}}, {"node": {"oid": "74fe45bf09636de6b1cf7d9135d0688ab0629c9b", "messageHeadline": "activate past_present_share_buffer for sampling node (#14166)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "be879c11ee38e34f4df2e6510a39641ee701f9e4", "messageHeadline": "Add batched and strided batched gemm as TunableOp (#13841)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 1642, "deletions": 256}}, {"node": {"oid": "5eac2c1f41a0e522c8ce70f5563a317c239e4c1d", "messageHeadline": "relational attention bias cuda op (#14149)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 485, "deletions": 0}}, {"node": {"oid": "8e2163018dae3e3f938e7206483f37ec423a2c81", "messageHeadline": "Ignore more build directories and clangd files (#14154)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "2cacb24cb02a3c4bd58689d547f8fd2332acc93c", "messageHeadline": "Add CrossAttention operator (#14146)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2396, "deletions": 2010}}, {"node": {"oid": "c6ff5bac9d36f61236bc896bfc9dde5f27d67e1b", "messageHeadline": "Update torch in eager mode CI pipeline (#14094)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 87, "deletions": 34}}, {"node": {"oid": "c65a03699a3790011ca8fdfeb8a2628f7aa280c9", "messageHeadline": "add ThreadingOptions, wraps OrtThreadingOptions (#13711)", "author": {"name": "we1559", "email": "zengxiangneng@gmail.com", "user": {"login": "we1559"}}, "additions": 70, "deletions": 0}}, {"node": {"oid": "babc1323e38c9ee862c43284849ec979bd3b5964", "messageHeadline": "Consolidate Identical Children Nodes (#14026)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 257, "deletions": 3}}, {"node": {"oid": "d0c5ffd5f7c165cf171e900c58454358d21032ee", "messageHeadline": "Misc transformer fixes - 2 (#14156)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 201, "deletions": 98}}, {"node": {"oid": "37028066531dbe60bfc36b7c28f5512c26011c5f", "messageHeadline": "[ROCm] add softmax, topk, layernorm to microbench (#13997)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 266, "deletions": 0}}, {"node": {"oid": "b222a8e01bdf7537c3ac633fa23e45e62150f14d", "messageHeadline": "[Fix] build error with MIGraphX tag rocm-5.4.0 (#14141)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "0ed7277bbe5d865e136d5b41a94bbcc5e07d4ea4", "messageHeadline": "fix training compilation option (#14151)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2ce7b1c1dcc2e21a79be672d2fcf2c480fafd8b0", "messageHeadline": "Enable cache for msbuild (#14085)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 134, "deletions": 18}}, {"node": {"oid": "d460c01b8c33dcd13381180348ef95b440a54946", "messageHeadline": "Fix skew between GPU/CPU timestamps in ORT profiler (#14004)", "author": {"name": "Abhishek Udupa", "email": "abudup@microsoft.com", "user": {"login": "abudup"}}, "additions": 54, "deletions": 23}}, {"node": {"oid": "90cff21fa72bd103b3666fc3a72c7ba7504acebe", "messageHeadline": "Avoid the lock for device stream impact the cpu build (#14131)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 127, "deletions": 83}}, {"node": {"oid": "4eac0db3afecf93063bf3d4ff289763453aaf89f", "messageHeadline": "[ROCm] Add GemmFastGelu CK implementation (#13759)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 857, "deletions": 385}}, {"node": {"oid": "2b45410e52e55dd16ac45802f8683c8bd64de6c2", "messageHeadline": "Fix Prefast warning in CUDA contrib op (#14074)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "68794d0ac168a162c0e74e1b135a600aa046b057", "messageHeadline": "Improve custom op library handle cleanup (#14099)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 506, "deletions": 230}}, {"node": {"oid": "dc997af69536ed27cbd7987082cd162fc1c88661", "messageHeadline": "Use RegisterOp to add Op instead of directly manipulate base class fi\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 45, "deletions": 39}}, {"node": {"oid": "b313055ad6301637bf4b7e1fb114d82f940cf62d", "messageHeadline": "Updated issue router to migrated project (#14114)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ae148ebc05cd1a5afd1f0c25de15eb561c1c0794", "messageHeadline": "T5 skip_layer_norm cuda op (#14093)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 256, "deletions": 57}}, {"node": {"oid": "b6ea60436d2bf6fa5896652a8d0caa55bc3cd1a2", "messageHeadline": "[DML EP] Decouple the bucketized allocator from the individual block \u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 148, "deletions": 54}}, {"node": {"oid": "f344d4b3d196fdea5a2c19a33715ef2a64590f7a", "messageHeadline": "Label issues with mobile when android or ios are present (#14033)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "821baa5b839b04581ab9b7a8b8cd44bb5b002c5a", "messageHeadline": "Support generation script with custom eos/pad token id (#14113)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 21, "deletions": 0}}, {"node": {"oid": "e5e3570ac50001566584607913b4286383fcda15", "messageHeadline": "fix cg issue (#14112)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d279ce2f675ac2f99f854c09c7d7fcef2944eff5", "messageHeadline": "bug fix, NNAPI filter out un-supported graph (#14040)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 38, "deletions": 4}}, {"node": {"oid": "15c1157ef2d179c911f4e16f87f074ea55ea8c8b", "messageHeadline": "New Pattern Support for LayerNormFusion (#14118)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 91, "deletions": 17}}, {"node": {"oid": "f864b543934b2d45fe0e1180f723ba61d9be047b", "messageHeadline": "Use today's cache only (#14120)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "bdeba4e31c0ed2039007821e2326f8c8b344d10d", "messageHeadline": "Bump json5 from 1.0.1 to 1.0.2 in /js (#14109)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 13, "deletions": 19}}, {"node": {"oid": "0ff61f7b971b98b93f75d4bb7bbb81fb4553a73c", "messageHeadline": "Update torch to 1.13.1 in CI and packaging pipelines for ort training\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 32, "deletions": 300}}, {"node": {"oid": "b29a1c73485beaa37771d8996284ba575f8f065e", "messageHeadline": "Address follow-up comments on multistream pr #13495 (#13992)", "author": {"name": "cao lei", "email": "jslhcl@gmail.com", "user": {"login": "jslhcl"}}, "additions": 178, "deletions": 158}}, {"node": {"oid": "a78ab4fbef513ece0c407e1360fa4e5e1f20b46a", "messageHeadline": "Add labeled to docs workflow trigger (#13979)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "68b5b2d7d33b6aa2d2b5cf8d89befb4a76e8e7d8", "messageHeadline": "Refactor training build options (#13964)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 190, "deletions": 143}}, {"node": {"oid": "d43e0ec9ba75db16c7d151f3281a61ae4168777e", "messageHeadline": "Misc transformer fixes (#14103)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 35, "deletions": 11}}, {"node": {"oid": "e7f9d40dde37960e0a4bfa9e98e29d6a37d8435f", "messageHeadline": "[DML EP] Force instance norm inputs to be 4D to better target metacom\u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 17, "deletions": 5}}, {"node": {"oid": "589612106aea3033d4f574cd194a532ec421bd46", "messageHeadline": "[DML EP] Force layer norm inputs to be 4D to better target metacomman\u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 25, "deletions": 4}}, {"node": {"oid": "587e891cae05e9d58f61a508c2ffeb95fcb146d6", "messageHeadline": "CloudEP (#13855)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1231, "deletions": 108}}, {"node": {"oid": "52e3fe961dda1aa769cd0b2b2551845f8a54169f", "messageHeadline": "add dnnl dependency in unittest.cmake (#14104)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 11, "deletions": 7}}, {"node": {"oid": "613920d6c5f53a8e5e647c5f1dcdecb0a8beef31", "messageHeadline": "Remove all usages of hipLaunchKernelGGL (#14089)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 84, "deletions": 84}}, {"node": {"oid": "6a9dc6c993a89b6b0014d9ba8f33da1fd0599083", "messageHeadline": "[CUDA] Update fused MHA to support flash attention and causal mask (#\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2105041, "deletions": 47345}}, {"node": {"oid": "5d729839b5d00cdbe13d59b4c8c481ca25f401b3", "messageHeadline": "Support loading widechar paths on windows (#14066)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 45, "deletions": 33}}, {"node": {"oid": "b85878953f8bae6d2d9afd48c843c1b76105a913", "messageHeadline": "Fix nightly ort training ci pipeline (#14007)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "9f52a8bc55a77b2c17f519dce803273c55a5efe0", "messageHeadline": "fix the cuda EP test failure (#14087)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "a79d88ed7f856836a7ff3d35e4277489c4bbdd2e", "messageHeadline": "Fix prefast scan bug 8995 (#14077)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "0c3480e56551cbb66339d765fc7c314890200ce2", "messageHeadline": "[ORTModule] ATen upsample_nearest Gradient Bugfix (#14069)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 35, "deletions": 25}}, {"node": {"oid": "770fb9649b338300303217dd11eb59b8683a469a", "messageHeadline": "[Fix] CPU memory type need device_id=0 to get allocator (#14050)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "d762aa2a4cdd1e2c08de9c0766879776c250e7a9", "messageHeadline": "Let Cmake decide where to place abseil (#14057)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "3bbcc2799fe19bfb26962851d7b3b6002a91d323", "messageHeadline": "Support for custom op variadic inputs/outputs (#13946)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 701, "deletions": 24}}, {"node": {"oid": "8ac264b89682ddc576e2eba50798eb49f0a377bb", "messageHeadline": "Deprecate one step beam search (#14046)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 45, "deletions": 2259}}, {"node": {"oid": "e49f358686f7f8ab16e49e4989c5de15d359afa8", "messageHeadline": "expose lr scheduler python bindings for on device training. (#13882)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 119, "deletions": 23}}, {"node": {"oid": "b999022b03a7cd88ecafb9ff804d32039aeb405c", "messageHeadline": "disable some test input for generation model temporarily  (#14045)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "68518a1b72e80b6d081fee496b5ec320a656be61", "messageHeadline": "Sampling op (#13426)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2404, "deletions": 509}}, {"node": {"oid": "4d2dc8bbbdcbbec6db89c3da51f997d89d8d04ba", "messageHeadline": "Replace all numpy.bool by python builtin bool (#14014)", "author": {"name": "fxmarty", "email": "9808326+fxmarty@users.noreply.github.com", "user": {"login": "fxmarty"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "1b58331fb342d7db457df1b7a735d23619d51927", "messageHeadline": "[QAT] Graph transformer to fuse QDQ pattern into FakeQuant (#13777)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 307, "deletions": 4}}, {"node": {"oid": "944bff0ad6497d8cf8bef84c10f45f9ef1c2025d", "messageHeadline": "Support two stages onnx GPT-2 conversion (#14025)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 369, "deletions": 165}}, {"node": {"oid": "694ba033e91f71d3585a92c6e48d09e7f3181083", "messageHeadline": "[ROCm] update skip_layernorm test sample (#14051)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2f5bf75e512ba37b6729e17f7bfba474a5836d3c", "messageHeadline": "Optimize computation orders (#13672)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 3802, "deletions": 547}}, {"node": {"oid": "7ed8bd4f95977d7601250564e3f40039d176a2c7", "messageHeadline": "Support (Bias)SkipLayerNormalization fusion in GPT2 (#13988)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 546, "deletions": 169}}, {"node": {"oid": "baba312e30fcea6c295c6add7492389045ba0b82", "messageHeadline": "Add provider selection for gpt2/convert_to_onnx.py (#13982)", "author": {"name": "Joseph Groenenboom", "email": "joseph.groenenboom@amd.com", "user": {"login": "groenenboomj"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "a170e40fbb1a801fcb446e2b2fa28d3b2cb151d3", "messageHeadline": "[ROCm] Update Dockerfiles of ROCm and MIgraphX to ROCm5.4 (#14013)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 13, "deletions": 9}}, {"node": {"oid": "b5fd2a6a80ab429c50dbf3078966aaccc79e7fdb", "messageHeadline": "[ROCm] Add ROCm5.4 to python package pipeline (#14012)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "ab2dd8dfaf4cd2eb2eb52e991cc59b2d173cf835", "messageHeadline": "[ROCm] Update ROCm and MigraphX CI to ROCm5.4 (#14011)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "df8ff34f25cbff4f15f459210822668d1237cc0b", "messageHeadline": "Update CUDA ArgMin/ArgMax op kernels to have end version 11 since ops\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 36, "deletions": 28}}, {"node": {"oid": "8943d623a4f3e85b629761dd560b2a7971d7b885", "messageHeadline": "DML EP Register operators for Opset 16 (#14034)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 24, "deletions": 5}}, {"node": {"oid": "1a177a17133983cf736b0628fb1d90e9db335f2a", "messageHeadline": "Cover beta in all Conv paths. (#14008)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 81, "deletions": 18}}, {"node": {"oid": "ccc44875535fd78eee74911bac4bf37abf20e65e", "messageHeadline": "fix CI onnxruntime_test_python_sparse_matmul.py (#14039)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "7738be9b256f31f5328035d6ddd3cae87b29ae72", "messageHeadline": "[prefast:Warning]: C26451 (#14036)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "05137e6ec44b66367517ff6989c9251e4f5c1e6f", "messageHeadline": "Use target name for flatbuffers (#13991)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 79, "deletions": 64}}, {"node": {"oid": "cd305a90d61576296b126dbf2f9a098f0f255e3e", "messageHeadline": "Stop creating static thread pool to fix random hang in onnx_test_runn\u2026", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 12, "deletions": 15}}, {"node": {"oid": "533fe37cbd6be402e0975e35ac7c26fe3dde891b", "messageHeadline": "fix build break in transformer debug dump (#14009)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "fc2a6db573689f72e36b2e0dfaccef000096a35b", "messageHeadline": "Update absl to the latest release (#13990)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 63, "deletions": 111}}, {"node": {"oid": "f1044e3b9a9cb5b0db0b72a2d5ead4c62d76a278", "messageHeadline": "CUDA GreedySearch ProcessLogits optimization (#13823)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 104, "deletions": 64}}, {"node": {"oid": "28e2b1790fbf85f53c1b4386e7bbc00128ac1a40", "messageHeadline": "Moving MLAS threaded QGEMM packing buffer from stack to heap (#14002)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 118, "deletions": 13}}, {"node": {"oid": "fba09faf5be060b29b3dd0aea079a11c98c2da2b", "messageHeadline": "Implement reuse past and present tensor in Attention Ops. (#13791)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 889, "deletions": 227}}, {"node": {"oid": "2df046fc67fdb0a5fcb2be892e8a5aa88d5e52b4", "messageHeadline": "Fix deprecated-builtins (#14001)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 66, "deletions": 65}}, {"node": {"oid": "6fb54fc607d94de50a18997430d7c87b0d921e87", "messageHeadline": "Add ms domain during saving onnx model in onnx_model.py (#13978)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "cc0a6213e4dfba3c2bc1d684e5b4d3054242e5f0", "messageHeadline": "[js] update versions of a few build dependencies (#13977)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1384, "deletions": 1980}}, {"node": {"oid": "ba89cae3bdfe344d2c54780945995de07fbd9099", "messageHeadline": "Update package pipelines to support TRT 8.5  (#13998)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "848f80f7a9a910283765a57bd4b3666cd6bd6084", "messageHeadline": "Skip some attention op tests in A100 (#13980)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 15, "deletions": 0}}, {"node": {"oid": "6705915af87827f4bfd0ed672134f0c0df237d45", "messageHeadline": "[CANN] Add the ability to run graph (#13728)", "author": {"name": "FFFrog", "email": "85333600+FFFrog@users.noreply.github.com", "user": {"login": "FFFrog"}}, "additions": 718, "deletions": 55}}, {"node": {"oid": "aa9fbed3d4172f9d2c3c7c61ca1b8372b0cdaa86", "messageHeadline": "Add compilation cache for Linux GPU (#13995)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 55, "deletions": 16}}, {"node": {"oid": "be9ae28d9f5e39a85b93e83775513649540f5279", "messageHeadline": "Add ability to set RunOptions config entries to C# API. (#13939)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 39, "deletions": 2}}, {"node": {"oid": "7d20d889d1b4924f9b0a99deb7f1fa0adfac2448", "messageHeadline": "Use cache for compilation in container (#13960)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 44, "deletions": 5}}, {"node": {"oid": "a061fedb5dede8d86b5d8757abb3faaa1dcd299b", "messageHeadline": "Exclude affinity-setting logic from minimal build (#13967)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 23, "deletions": 2}}, {"node": {"oid": "0ee5a5f229a30197e9fcc4236c89cdc5f91888a6", "messageHeadline": "[debug] allow dump node placement in transformer models (#13954)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 18, "deletions": 0}}, {"node": {"oid": "c4ecbb96d9b9432df39638dc1f0eda55e3e05f35", "messageHeadline": "Fix issues in TRT model ID generator (#13837)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 98, "deletions": 436}}, {"node": {"oid": "b52e8bf718e59b621438c747960f5e8d3019a927", "messageHeadline": "[oneDNN ep] QAttention BF16 and GPU support added (#13793)", "author": {"name": "Sunny Shukla", "email": "102830081+sunnyshu-intel@users.noreply.github.com", "user": {"login": "sunnyshu-intel"}}, "additions": 132, "deletions": 36}}, {"node": {"oid": "c882601425a0c4958842503355ed0e281546858a", "messageHeadline": "Add noexcept annotation to address prefast warnings (#13965)", "author": {"name": "Abhishek Udupa", "email": "abudup@microsoft.com", "user": {"login": "abudup"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "a3cd36dbfbc9b3f33e5abef852752ac4e5472198", "messageHeadline": "change default cudnn_conv_use_max_workspace =1 (#13981)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "a81faee41ef2344de448caecb0f42a34fdc9ead7", "messageHeadline": "Multi-stream execution support (#13495)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 8265, "deletions": 5193}}, {"node": {"oid": "f4cd35f9b1301f54d65a3e59c525b92e85bf384e", "messageHeadline": "[xnnpack-ep] NEW EP API in objc (#13941)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 124, "deletions": 0}}, {"node": {"oid": "a9b1fb032bbbd98f365adbbff77b32530d5dde20", "messageHeadline": "FIX: macOS CI pipeline doesn't run tests (#13970)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "1fd63487fd8f62681f385e9c3e42fe1574735492", "messageHeadline": "ORTModule support for kwargs input that is a dict (#13910)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 116, "deletions": 50}}, {"node": {"oid": "3b17ab7c6508f3d336cc99c475c389f188cd4631", "messageHeadline": "Add float64 kernels for Floor, Ceil, IsNaN (#13906)", "author": {"name": "Jakub Bachurski", "email": "kbachurski@gmail.com", "user": {"login": "jbachurski"}}, "additions": 70, "deletions": 11}}, {"node": {"oid": "5a55fac402284c1cb5e7544395ff5f70cd165506", "messageHeadline": "Miscellaneous updates to training apis (#13929)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 470, "deletions": 280}}, {"node": {"oid": "e5f6689ae75b7f3c1161bbf8cfc07b392f5cd25a", "messageHeadline": "Allow Tensor to be scalar if it is not per channel. (#13959)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "5b492cbae3c4c94b93168cff95e924776f4205f7", "messageHeadline": "[TensorRT EP] support TensorRT 8.5 (#13867)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 271, "deletions": 52}}, {"node": {"oid": "8c249cc8f7933fa031e608bf2e10a03ae93dfa31", "messageHeadline": "[QAT] FakeQuantGrad and gradient building for FakeQuant (#13825)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 266, "deletions": 5}}, {"node": {"oid": "6090d8cd6eb89ab8da000ba41ccabb48ce798feb", "messageHeadline": "Fix usage of enable_training_ops and reduce ifdef complexity for trai\u2026", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 160, "deletions": 118}}, {"node": {"oid": "7894d44d2dc18cf3b7eac480c6040fd1905a3478", "messageHeadline": "Improve MacOS Cache Code (#13958)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "6900109ee88ae7262a65495e10d8558c4c24b84d", "messageHeadline": "Bugfix for GetCpuPreferredNodes (#13590)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 13, "deletions": 2}}, {"node": {"oid": "80a046b36f6ff82071976e58fe525a587bb2df96", "messageHeadline": "[ROCm] update amd CI huggingface model performance number (#13961)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4b54b9d5b07df5270c5dfb31aefe12f881796b75", "messageHeadline": "[ROCm] Sort kernel explorer profile result (#13862)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 218, "deletions": 84}}, {"node": {"oid": "abc5c25a85c13dbd310b74de8ba26b2f139e7396", "messageHeadline": "Updates to GreedySearch/BeamSearch (#13943)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 438, "deletions": 59}}, {"node": {"oid": "b4dd5dda1208c6329bf96d0f804cb85e0664c7a4", "messageHeadline": "Revert \"Update protobuf version to 3.18.3 in tools/ci_build/github/li\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c9edc01c0b02f2d31020ac4bc06631f755403d3b", "messageHeadline": "[ROCm] float16.h should use __HIP__ not USE_ROCM (#13684)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "9dd593e18ffb3ca6aa0e6a325206a4546d0abe77", "messageHeadline": "[DML EP] Fix prefast warning in LayerNorm ops (#13889)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "8246ff015a7b44bb60e89d88181d13527f72eb1e", "messageHeadline": "[DML EP] Add EmbedLayerNorm (#13868)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 551, "deletions": 4}}, {"node": {"oid": "e5ed47a11dd53bf87791d137307c2a898184087f", "messageHeadline": "Bug Fix - ORT Web build script (#13925)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3addbabc59d63ff12d7a5136e9470124929213a3", "messageHeadline": "Fix react native ci (#13948)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "9836a4ed1e676b1839a909326020b0eaac43dba9", "messageHeadline": "Bump engine.io and socket.io in /js/web (#13723)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 64, "deletions": 137}}, {"node": {"oid": "067d42530678d2e406eb673c730e983d0a749f1a", "messageHeadline": "Correct logic for GPU backend detection (#13944)", "author": {"name": "Joseph Groenenboom", "email": "joseph.groenenboom@amd.com", "user": {"login": "groenenboomj"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "526368f4d81651829ee63e8ee334ad2a292e46d4", "messageHeadline": "Auto add docs issues to project (#13897)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 17, "deletions": 0}}, {"node": {"oid": "69fdb7c75a83000ddf1afce1ebc705c2adb2e476", "messageHeadline": "Labeler: Add ROCm ep issue labeling (#13923)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "928c44f71b5d255dcb714779bed554e991bd7fbe", "messageHeadline": "fix(cuda): install missing python3-packaging in Dockerfile", "author": {"name": "Jian Zeng", "email": "anonymousknight96@gmail.com", "user": {"login": "knight42"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d7d932c1c223d8749f1962d5bab8af2fcee5b634", "messageHeadline": "Cjian/where python operator (#12795)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1056, "deletions": 26}}, {"node": {"oid": "b23395f9774653dc0217d98de48b1331f2a7c5ae", "messageHeadline": "Update protobuf version to 3.18.3 in tools/ci_build/github/linux/dock\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "bd189d56ad68effd47ea0c7ada23f37069a12b06", "messageHeadline": "Bump decode-uri-component from 0.2.0 to 0.2.2 in /js/react_native (#1\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "070769d61dbc0b0dda5fae5ace3478c19ec064eb", "messageHeadline": "Use onnxruntime_fetchcontent_makeavailable cmake function for TRT (#1\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "75584c5fa8d71f6acd2f78af11f279ecbdc47fc1", "messageHeadline": "Enabling thread pool to be numa-aware (#13778)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1178, "deletions": 607}}, {"node": {"oid": "b8d941f065f36fba47d2870266a70aa05dec36f4", "messageHeadline": "Cjian/pad ops bug (#13930)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 83, "deletions": 14}}, {"node": {"oid": "fe827c3891b2ce38e278fe5cbd9de5ee821eae50", "messageHeadline": "[DML EP] Disable DML Graph Fusion for lower graph optimization level \u2026", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 30, "deletions": 18}}, {"node": {"oid": "8cfbc4fe9112c173e4b86bada2467a155bdd3513", "messageHeadline": "Add support for other data types to Split CPU kernel. (#13900)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 135, "deletions": 184}}, {"node": {"oid": "2cb12caf9317f1ded37f6db125cb03ba99320c40", "messageHeadline": "Output cache stats (#13937)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "89812a623e467000e85ad116a146820fe2e69947", "messageHeadline": "Add two daily build jobs to validate some extra build configs (#13921)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 35, "deletions": 1}}, {"node": {"oid": "51aaf2e021de9a792c0ece6e02d61248a79b763d", "messageHeadline": "Allow using separate GPT2 decoder subgraphs for the initial run and t\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 286, "deletions": 8}}, {"node": {"oid": "22fa62152a14520086ff4088eed0b446954c5e8e", "messageHeadline": "Pass SessionOptions to XnnpackProviderFactoryCreator. (#13318)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 44, "deletions": 19}}, {"node": {"oid": "87eef1fe21a81b62ffee333504a12d98de47e5a8", "messageHeadline": "Use updated ONNX license in ThirdPartyNotices.txt. (#13919)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 198, "deletions": 9}}, {"node": {"oid": "a7bc927b4b1f4234b44f902b213db09fde0e8d3b", "messageHeadline": "fix typos in training apis (#13908)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "1921d8463608a9f6a51ac37d743f580cef71e2c0", "messageHeadline": "Updated ORT-Web build instructions (#13282)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 1, "deletions": 58}}, {"node": {"oid": "21dd341e52edbeb9fb1598c537e612065aec4d67", "messageHeadline": "Add Google Analytics to python apidocs (#13901)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 9, "deletions": 2}}, {"node": {"oid": "db9c677b6318fdd96c8c8aa23d9bca8a6c4b345e", "messageHeadline": "[EP Perf Dashboard] Add TensorRT 8.5.1.1 dockerfile (#13843)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 84, "deletions": 0}}, {"node": {"oid": "83c59d259456773a01434767ac5925322001619c", "messageHeadline": "Session-aware and thread-safe CUDA profiler (#13706)", "author": {"name": "Abhishek Udupa", "email": "abudup@microsoft.com", "user": {"login": "abudup"}}, "additions": 782, "deletions": 632}}, {"node": {"oid": "18d5cd6ee5dee1e3e2d43372b8681bd0b209309e", "messageHeadline": "Bump Newtonsoft.Json from 13.0.1 to 13.0.2 in /csharp/test/Microsoft.\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6f25f6e1f0d089aec1a8263b81f066b38d0c92a7", "messageHeadline": "Bump Newtonsoft.Json from 13.0.1 to 13.0.2 in /csharp/test/Microsoft.\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fba49952d208d192d5aa9254d8f2bab7dd4d05bb", "messageHeadline": "Bump Newtonsoft.Json from 13.0.1 to 13.0.2 in /csharp/test/Microsoft.\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d5b45226bee1f86c0e1a483ffb6473cad7d3b066", "messageHeadline": "Improve the handling of /external:I (#13904)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "d8e22f6e50182b89a6b1c24b17676462f25091a5", "messageHeadline": "Update VerifyOutputs() to use SpanEq() instead of gsl::span compariso\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dead5c6b3a5030c40238b51d58ccc003dc09b2e9", "messageHeadline": "Revert \"[js/rn] support load model from buffer on Android (#12676)\" (\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 38, "deletions": 99}}, {"node": {"oid": "d22be84add1e5eaff2bda245cb77b87f5473588f", "messageHeadline": "Pin packaging to version 21.3 to address training pipeline failures", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "05dc1165a5e03d35cfe573c5a7d0dff7bcb22da5", "messageHeadline": "Add protobuf version constraint (#13870)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fb4707f76d9161b3f8fdc4ba294e3533382f30f3", "messageHeadline": "add cuda support to python bindings (#13700)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 68, "deletions": 13}}, {"node": {"oid": "7d684d125559685bc4a6c32248cb5e60f04d84af", "messageHeadline": "Include algorithm selection exposed by ROCBLAS extensions API in GEMM\u2026", "author": {"name": "Abhishek Udupa", "email": "abudup@microsoft.com", "user": {"login": "abudup"}}, "additions": 234, "deletions": 12}}, {"node": {"oid": "dbf47284d1769a5f57b3c7b51a6806168a422274", "messageHeadline": "[wasm] disable closure compiler in debug build (#13865)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "81c2defd3b7081041db7ca27e3a28351cf039083", "messageHeadline": "Remove unused git submodules (#13830)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 575, "deletions": 137}}, {"node": {"oid": "c1cc1d5859da460f2320389252da3690dd2dfbbc", "messageHeadline": "[ROCm] Update FastGelu and add kernel expolrer test for FastGeluStati\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 112, "deletions": 60}}, {"node": {"oid": "7694b695a91170f9ccf30f14f184b05d3e2dce66", "messageHeadline": "[ROCm] Simplify ROCm manylinux dockerfile (#13873)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 26, "deletions": 511}}, {"node": {"oid": "a64ddb36d0177a17d36f4920b0ca2177966950e3", "messageHeadline": "Always build with XNNPACK EP in iOS CI build. (#13850)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 70}}, {"node": {"oid": "5b16593192e03b728f8b096e04a6a79f802670e0", "messageHeadline": "[DML EP] Attention Kernel bug fix (#13879)", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "4c79977f5277885424635ab796108e4dee757d82", "messageHeadline": "[wasm] fix session option setting of mem_pattern (#13858)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ffdcde7cc71b596330dce6ad45ce6f4ac834a1bf", "messageHeadline": "Bump minimatch from 3.0.4 to 3.0.5 in /js/web (#13722)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 27, "deletions": 6}}, {"node": {"oid": "f453d2845e7d68f425fcce1a9db872ebf9dc71d7", "messageHeadline": "adding get and set lr for optimizer (#13661)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 44, "deletions": 0}}, {"node": {"oid": "983877c7129957541f9d5b49943bf83104868616", "messageHeadline": "Decouple strided tensor support from ENABLE_TRAINING (#13829)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 50, "deletions": 49}}, {"node": {"oid": "f6c493793dfa85ef889ae5400d0c9caa268dd38c", "messageHeadline": "Revert \"skip TestCUDAProviderOptions in End2EndTest (#13737)\" (#13874)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ae2a9373ab93ea0def447e4ff567bbbf8ad4d8d7", "messageHeadline": "reenable quant model tests (#13871)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 0, "deletions": 14}}, {"node": {"oid": "96d8d2c2780c395055dcbcb03cd58cc1f2f393ac", "messageHeadline": "[DML EP] Add SkipLayerNormalization (#13849)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 179, "deletions": 4}}, {"node": {"oid": "004a1538d365a3a673ba51aeadb6ecec77e224e0", "messageHeadline": "Extend vocab padding for logits MatMul for fp16 GPT2 GreedySearch (#1\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 105, "deletions": 6}}, {"node": {"oid": "f79d38181b0a3e0fab7fb1ea829be2467163a671", "messageHeadline": "Fix hipify to avoid nccl_service.h: No such file or directory (#13852)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "d12521d7b2f46d5aeda1306e5d8243df032f1344", "messageHeadline": "Upgrade pybind11 (#13853)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "78d18fbf34a5a631124aa258e7e9c34a61940563", "messageHeadline": "Use CacheTask to Accelerate MacOS build (#13859)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 23, "deletions": 1}}, {"node": {"oid": "d2188fbff99a8f06982d902e89e6de139d4a627e", "messageHeadline": "skip resnet50-int8 model test in training (#13856)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "65201e47bf58ce5bbc8389a83082f69ee1ea50b3", "messageHeadline": "Enable nuget packages for on device training (#13637)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 127, "deletions": 61}}, {"node": {"oid": "d5574e6999185daeba7f3e0b7c60d3fd3dbcbb8d", "messageHeadline": "LayerNorm test fix (#13840)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 19, "deletions": 12}}, {"node": {"oid": "5f4e0c95ec1ea21baa788102f672a949e1b3f8b4", "messageHeadline": "Misc minor bug fixes in transformer kernels (#13780)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "f34ebbc8ffdb077151f6a4d9b743f34dbda262a2", "messageHeadline": "fix a wrong assert condition in benchmark_helper (#13821)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "335b62bde6cfc5fa88cb199ad8f42aeb400857c0", "messageHeadline": "Fix invocation of GetInputMemoryType. (#13828)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "b53bbe7370a0270883d1ecb11be3a3d282ba4f17", "messageHeadline": "[DML EP] Add an implementation for NonZero (#13768)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 205, "deletions": 3}}, {"node": {"oid": "b9702587dffa0604924a7a6a10481f54156032c7", "messageHeadline": "[oneDNN] Implemented Concat Op (#13646)", "author": {"name": "Gaz Iqbal", "email": "gaziqbal@gmail.com", "user": {"login": "gaziqbal"}}, "additions": 230, "deletions": 6}}, {"node": {"oid": "c2d08fd73af3f1b1ca414f3c2bc83827e3909f7d", "messageHeadline": "[DML EP] Add support for LayerNorm (scale == nullptr) != (bias == nul\u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 20, "deletions": 47}}, {"node": {"oid": "a0b470bc35cb2df26d3dc8c75c5261f92e8cf586", "messageHeadline": "[DML EP] Add mixed datatype support for DML's LayerNorm contrib op (#\u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 286, "deletions": 18}}, {"node": {"oid": "82d123b6c91845c7363981963ae3ba0843c5c4a5", "messageHeadline": "[quick fix] Build onnxruntime under DISABLE_ABSEIL (#13799)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "04900f96c1e22f99f8cc12b7ee98930f03cd93f9", "messageHeadline": "Improve dependency management  (#13523)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1841, "deletions": 2025}}, {"node": {"oid": "e9b92fdf33cd6bbd70ee54ac395bd1c58d6c93b6", "messageHeadline": "[DML EP] Add DML implementation for BiasGelu (#13795)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 53, "deletions": 1}}, {"node": {"oid": "e0dcbc3832850805fd60d5f38cdb810d8c3cc80e", "messageHeadline": "Fix C26436 prefast errors (#13774)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "4128e44b4f7aa87ca0b907c906aab63bde7219c3", "messageHeadline": "[DML EP] Upgrade DML to 1.10.0 (#13796)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "777c474f616d486ae2f7895980ecfa88a0b8373d", "messageHeadline": "skip quantized model C# tests on GPU (#13782)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 9}}, {"node": {"oid": "7df8f84228020f3095d4600d027912a94ece30d4", "messageHeadline": "Improve DORT document (#13790)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 62, "deletions": 9}}, {"node": {"oid": "77c97b6f168d529a75fe39c4948e6baaa561fc80", "messageHeadline": "[js/rn] support load model from buffer on Android (#12676)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 99, "deletions": 38}}, {"node": {"oid": "639d285670c7f5dcbc957ecbce350c295cbe328a", "messageHeadline": "[DORT] Catch up with yesterday's PyTorch change (#13779)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 13, "deletions": 38}}, {"node": {"oid": "441b30b2d26d36ca1db2930ade2fe82622ce0cd4", "messageHeadline": "Move a function call outside a loop in ORTModule (#13771)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "08ed09d20bd8670bfa84e8dd3c113fce02ec4008", "messageHeadline": "Add DML support to the transformers benchmark.py script (#13776)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "29ed8811e5946bb710a9b1915641214b9ac628e2", "messageHeadline": "Move C/C++ deps' URLs to deps.txt (#13769)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 400, "deletions": 15}}, {"node": {"oid": "571dc5a1f1ab35a5159a558a1ca7e1171608d963", "messageHeadline": "Support exteranl weights in DML execution provider (#13740)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 48, "deletions": 35}}, {"node": {"oid": "ce0025d3f27da0a60155a7b473563f736919120c", "messageHeadline": "Fallback Pow op in layer norm to FP32 in TRT to avoid overflow (#13639)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 63, "deletions": 5}}, {"node": {"oid": "0327606d2de6ad100b0bcae58d01024a97afe383", "messageHeadline": "Revert TRT EP Linux CI to run unit tests in container (#13766)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 3, "deletions": 65}}, {"node": {"oid": "abe1642a0cd5a12039bbd634c117d1588dd309da", "messageHeadline": "Update fusion for distilbert accuracy test on SQuAD (#13748)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 148, "deletions": 47}}, {"node": {"oid": "181628ced1bc0a07337ea7485dab63c9d425d6d2", "messageHeadline": "[CANN] add more operators (#13578)", "author": {"name": "FFrog", "email": "85333600+FFFrog@users.noreply.github.com", "user": {"login": "FFFrog"}}, "additions": 2102, "deletions": 482}}, {"node": {"oid": "2c2993884604dba7b2409443bfb1b56ff8a8d97e", "messageHeadline": "[QAT] Introduce FakeQuant op (#13649)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 435, "deletions": 0}}, {"node": {"oid": "49c37689853e6f61585effd4ad367de8b3dc8492", "messageHeadline": "Enabled ops for DeBERTa model (#13690)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "7c53b6eee8587a0ff689e8bc0c399649dfa533b1", "messageHeadline": "Skip the tests of saving tensor in backward (#13767)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "3c258c878c9428f80b4f20942efe35f3c9603b28", "messageHeadline": "[CUDA] Optimize Slice Kernel (#13641)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 234, "deletions": 190}}, {"node": {"oid": "47780b7f3b55a2518b694be5eaf8c20119272432", "messageHeadline": "[XNNPACK] add more computation heavy ops (#13270)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1800, "deletions": 613}}, {"node": {"oid": "4fbe16e493f7d3269e85a952c3612ce7b8a6251f", "messageHeadline": "Ifdef cpuinfo code on platforms we do not set affinity (#13486)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 13, "deletions": 17}}, {"node": {"oid": "2d523c507ec0945a356137833db55980ba6d75f3", "messageHeadline": "for wasm catch exceptions at top level api (#13644)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 40, "deletions": 18}}, {"node": {"oid": "b7c3862330be41545850a83aacd0e8e0ec98f04d", "messageHeadline": "Update resource section in readme (#13724)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "b4a4fa5aac9e1fafe348f2c9d7d121ed28a35433", "messageHeadline": "Fix compile error with protobuf RepeatedIterator (#13731)", "author": {"name": "Jicheng Tang", "email": "tangjch15@gmail.com", "user": {"login": "tangjicheng46"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "aa1390e963a4570c00684877598c508c931fe9ff", "messageHeadline": "Fix Prefast Errors (#13675)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "c6bea4f02f5e5208bc3fa0ad070ac3466e6782a8", "messageHeadline": "Modify MIGraphX EP for Accuracy tests (#13455)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 29, "deletions": 19}}, {"node": {"oid": "4ca62b9ee827d243d1b1356126a033cf8710b25a", "messageHeadline": "fix build break in test/beam_search_topk.cc (#13739)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 8, "deletions": 6}}, {"node": {"oid": "47e7630378cf36939ec99348f110a7ec77463621", "messageHeadline": "[CUDA] Transpose3DImpl Supporting more Cases (#13611)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 71, "deletions": 49}}, {"node": {"oid": "87d5703b14b0123b63a7ecdd0f69505fd30421c1", "messageHeadline": "skip TestCUDAProviderOptions in End2EndTest (#13737)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c1bda4c1cce4d0b71f6d27833b0ad900560e7db2", "messageHeadline": "fix buffer overuse in addtofeed() (#13733)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "e306b44e98d93911476ce5f9207a7e31f6d17700", "messageHeadline": "Improve coverage of fused MHA in Attention (#13732)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 71, "deletions": 17}}, {"node": {"oid": "87e6a26c5d99db784b8df4d76a85fba6ac344eb4", "messageHeadline": "Enforce Prefast check in Windows CPU CI pipeline (#13735)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 35, "deletions": 25}}, {"node": {"oid": "9168e2573836099b841ab41121a6e91f48f45768", "messageHeadline": "Patch  eval_squad.py script for Python < 3.8 and multiple Execution P\u2026", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 20, "deletions": 12}}, {"node": {"oid": "977da6635bf2442e790c4ac193023203f2d14d5e", "messageHeadline": "[ROCm] Remove tuning options on transformerOptions (#13689)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 40, "deletions": 57}}, {"node": {"oid": "c43ce64795df8d0284c8b612693ff237439f25d3", "messageHeadline": "Beam search TopK improvement (#13594)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 770, "deletions": 83}}, {"node": {"oid": "7857f59d2be276719322cb6dd1d0db4445f8ce80", "messageHeadline": "Use sequences to create initial feeds for decoder subgraph (#13719)", "author": {"name": "apsonawane", "email": "111780983+apsonawane@users.noreply.github.com", "user": {"login": "apsonawane"}}, "additions": 33, "deletions": 9}}, {"node": {"oid": "fb85b31facb9fb3fc99c76f99c93ea8f06ada39b", "messageHeadline": "Remove protobuf pin from training requirements (#13695)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2bebe6189a0481695219aa8cf47ef6d04510d218", "messageHeadline": "set node schema when apply NHWC transformer (#13660)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 12, "deletions": 1}}, {"node": {"oid": "ce460f9cdbd09fe82f47b16d66b6c35e55c8a460", "messageHeadline": "[DML EP] Return device removal reason when D3D12 device gets removed \u2026", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 33, "deletions": 24}}, {"node": {"oid": "6c5333e1a71a6a4b8eb56d70d97bd559a4452dbe", "messageHeadline": "[DML EP] Enable more DML tests (#13726)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "dd2c031d954112ed15114ba582827cf630d385de", "messageHeadline": "[java] Sparse tensor support (#10653)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 2218, "deletions": 106}}, {"node": {"oid": "8b0e0f49274419a3e2cebd48be3ed9d2b4fee334", "messageHeadline": "Add RemovePadding and RestorePadding for BERT model (#13701)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1320, "deletions": 61}}, {"node": {"oid": "ba9a585fcc954d781b4654f9d83fbba513245cbb", "messageHeadline": "Fix the tensor save for backward release problem (#13679)", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 71, "deletions": 4}}, {"node": {"oid": "947aab0ae0d3908680714d3d85f1d06bd0c973b5", "messageHeadline": "Make HF converge with lighting native amp (#13616)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 150, "deletions": 11}}, {"node": {"oid": "67e46a873a5b6e50d437d6d2e1f09f04514b3848", "messageHeadline": "Add '-DCMAKE_OSX_ARCHITECTURES=x86_64;arm64' when build protobuf from\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "8f3c6ea0dfae30a2f23d7ba72a39d0fb63d62c69", "messageHeadline": "[ROCm] Add GemmFastGelu TunableOp (#13589)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 809, "deletions": 151}}, {"node": {"oid": "45a895cdc39932fffb0399a664a9914db53dfcf7", "messageHeadline": "[ROCm] Fix static TunableOp (#13668)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 82, "deletions": 28}}, {"node": {"oid": "f1b5e4f1c907b01c274b93537c3a7c76abaac7e7", "messageHeadline": "[js] [deps] upgrade @xmldom/xmldom@0.7.9 (#13705)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 67, "deletions": 37}}, {"node": {"oid": "307ad1413a0c304beff6c328f1fe6c2e7317ddeb", "messageHeadline": "[js/web] support 'pytorch_half_pixel' mode for WebGL kernel 'Resize' \u2026", "author": {"name": "Seungwon Jeong", "email": "jsw1295@gmail.com", "user": {"login": "ncianeo"}}, "additions": 27, "deletions": 4}}, {"node": {"oid": "3119381011b87dc8a02d2353272013969a4cc421", "messageHeadline": "ORT Web build script (#12643)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 146, "deletions": 0}}, {"node": {"oid": "a5c2047dd14f252673e6887000058c7d6290ab07", "messageHeadline": "Fix the remaining Prefast warnings in CPU EP (#13707)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 919, "deletions": 799}}, {"node": {"oid": "8de5381e84818e763bbae29f2fcb3b39751665fa", "messageHeadline": "Add IsSupported support to Op functor (#13692)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 79, "deletions": 2}}, {"node": {"oid": "4a2a857030fdb2dca72d7b2f5e77b15e7bd0ee45", "messageHeadline": "Bug Fix - WASM build break (#13699)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "da2bd3ad4df3edcf2d6b849bba4e4372d4d2d48f", "messageHeadline": "[ROCm] Build ROCm CI with Release config and enable kernel explorer t\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 27, "deletions": 18}}, {"node": {"oid": "84728761551245e5424f6d660a5b4161b9324ba8", "messageHeadline": "Bump socket.io-parser from 4.0.4 to 4.0.5 in /js/web (#13608)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "43a7b520e48f5ea93cdb773c1ee10ac665223ce6", "messageHeadline": "Convert label config to one line regexes (#13702)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 6, "deletions": 24}}, {"node": {"oid": "2d732e97292e58821afef6792d457b63fd864771", "messageHeadline": "[js] [deps] upgrade minimatch@3.1.2 (#13703)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "c7329e004d2067a90b991a752c3260c462f27060", "messageHeadline": "Improve fp16 performance of GPT-2's logits MatMul while using BeamSea\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 301, "deletions": 47}}, {"node": {"oid": "c358d64b0ee36d0a700128459950a26f8097117e", "messageHeadline": "Bump loader-utils from 2.0.0 to 2.0.4 in /js/web (#13666)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "4901987d1d19336a263fb296c8a7e8403397b40f", "messageHeadline": "Remove SafeInt dependency from Objective-C API. (#13698)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 12}}, {"node": {"oid": "3e9e5e9d6d6cd8265954b9289bb0288b397f1862", "messageHeadline": "Patch Protobuf and ONNX's cmake files and enforce BinSkim check (#13694)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 190, "deletions": 121}}, {"node": {"oid": "6160ba0692a68d08885cd96015739334aabb9367", "messageHeadline": "Fix aten::_to_copy in DORT (#13682)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 62, "deletions": 4}}, {"node": {"oid": "07812a2fa6967810b201479c5c4cbb10a6910692", "messageHeadline": "Fix UT Failure on AMD for ORTModule's Conv Test (#13688)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "7a57976d1a159108454501a5678e66b0ff39e1a8", "messageHeadline": "Make natvis files work better (#13665)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "38a74af45d45314cc05eed6ad9bf1c29f1eaea0b", "messageHeadline": "Support position_ids broadcasting in EmbedLayerNorm (#13677)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 128, "deletions": 33}}, {"node": {"oid": "abfdb63e31fa30b207e8384c164f924dda2993c3", "messageHeadline": "Update protobuf-java to version 3.21.7 (#13630)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "d5721b3464239802309975393c265fd274ce83eb", "messageHeadline": "Fix wrong import path in docs (#13680)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "a50877ac999fa3ad7087c17606837d40b493ab48", "messageHeadline": "[ROCm] Add ROCm5.3.2 to python package pipeline  (#13664)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 310, "deletions": 3}}, {"node": {"oid": "de77c60e6e81212937d4c6e6d2b1ae8830a86ee0", "messageHeadline": "[oneDNN ep] SLN performance improvement for bias (#13620)", "author": {"name": "Sunny Shukla", "email": "102830081+sunnyshu-intel@users.noreply.github.com", "user": {"login": "sunnyshu-intel"}}, "additions": 69, "deletions": 81}}, {"node": {"oid": "b731cf397d5ddc584c6af96af2f14221b93c85a3", "messageHeadline": "Make static analysis happy (#13655)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 50, "deletions": 64}}, {"node": {"oid": "116079749e66187b961284c8793acb05c6096344", "messageHeadline": "Fix Mac CI in Packaging pipeline (#13671)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "8442d9df2c979e74cfe3a9eb884645b680b2cfb0", "messageHeadline": "Cjian/c4244 round 6 (#13663)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 240, "deletions": 189}}, {"node": {"oid": "2efd2878abe95aef74bb6a5624a28b585f2636e4", "messageHeadline": "[rn] Add uint8 typedArray support for react native android (#13622)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 34, "deletions": 1}}, {"node": {"oid": "359091f64a3e03f6691d27d153e881aeec2f21f1", "messageHeadline": "XNNPACK - GEMM & MATMUL integration  (#13126)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 478, "deletions": 0}}, {"node": {"oid": "55fb790d88654f533dd670cc6475f557b7874d9e", "messageHeadline": "DML EP allow squeeze-13 axes to be empty (#13635)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "3201a1f84109fa3ccd42cd8d60ede119a054bd9f", "messageHeadline": "Cjian/c4244 round 5 (#13645)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 226, "deletions": 218}}, {"node": {"oid": "9c6c219949ecae6bd0f1d0982b7a1811bc6928ec", "messageHeadline": "Enable shape-sensitive analysis in ProfileExplorer for GPU kernels (#\u2026", "author": {"name": "Abhishek Udupa", "email": "abudup@microsoft.com", "user": {"login": "abudup"}}, "additions": 31, "deletions": 4}}, {"node": {"oid": "4cd8b4269a27cbc209346a517b7f4733beda7452", "messageHeadline": "ignore dirty state of submodule XNNPACK (#13648)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "9e649d1ac482357c111a72b288a5b7d5cedb3902", "messageHeadline": "Allow CUDA EP enable or disable TunableOp via session options and env\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 146, "deletions": 40}}, {"node": {"oid": "2490cf84c99407f1f41625cc82385f2181fd0dde", "messageHeadline": "[QLinearSoftmax]remove input_shape check in Ctor (#13489)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 144, "deletions": 121}}, {"node": {"oid": "ad31ac466ba24a76a12d5320a1453380ecff1a5c", "messageHeadline": "Delete cpu-esrp-pipeline.yml (#13623)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 265}}, {"node": {"oid": "b1169635cc19a078aa5f778878d23c453e0b47f6", "messageHeadline": "Ensure graph resolve occurs after free dimension is overridden (#13634)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "6f6560a7b98d5e3005a596a7aacf0fdad90004bd", "messageHeadline": "fix to reduce peak memory usage in ort-web (#13323)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 136, "deletions": 32}}, {"node": {"oid": "197191e58c126efa7a05cd4a5ca03e8863fa3747", "messageHeadline": "Update pylint config to include valid short names (#13631)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "f0ff2c5de99fb1a2ad36b962ac0e5f2271950c84", "messageHeadline": "Cjian/c4244 round 4 (#13632)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 156, "deletions": 153}}, {"node": {"oid": "369a8224094941e5ee717c2611fe5d8b952a3ca6", "messageHeadline": "Share TunableOp between CUDA and ROCM EP (#13560)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 840, "deletions": 116}}, {"node": {"oid": "78147c2d955f9e7db4c108e40526f61fbf9e7479", "messageHeadline": "Pin react-native version in js/react_native/android/build.gradle. (#1\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "9954454c65086c49b7c00f83b23ada76975f3546", "messageHeadline": "Make the ROCM profiler thread-safe, session-aware and preserve logica\u2026", "author": {"name": "Abhishek Udupa", "email": "abudup@microsoft.com", "user": {"login": "abudup"}}, "additions": 717, "deletions": 841}}, {"node": {"oid": "240a7ecf865c203892dde045f109b2c720638ef1", "messageHeadline": "Fix lgtm C++ error (#13613)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cd85a6333a24be1cc2dc1dd76dad5bf5cad60d55", "messageHeadline": "Add Missing Test File (#13607)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "31cb3cb254615a1414e7323a4da9ff0b1a0f727d", "messageHeadline": "[DML EP] Revert DML's cpu fallback logic (#13605)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 21, "deletions": 14}}, {"node": {"oid": "a89015b9407622e811140eac43a811beec1a0d55", "messageHeadline": "[XNNPACK] wraps xnnpack alloc with cpu_allocator (#13349)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 95, "deletions": 7}}, {"node": {"oid": "d286822464a23a530153488417cce8c5a094646b", "messageHeadline": "Fix round 4 (#13609)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 124, "deletions": 124}}, {"node": {"oid": "2bda3fd341ac0929653834d3565ba01d02342b7d", "messageHeadline": "Gather to Slice Fusion (#13599)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 218, "deletions": 4}}, {"node": {"oid": "05114437821871b5c9d14b22ea3477fd0726178a", "messageHeadline": "Cjian/c4244 round 3 (#13583)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 42, "deletions": 42}}, {"node": {"oid": "86968d1351badc151b829973a13414170c1db56b", "messageHeadline": "Merge win-gpu-ci.yml and win-cpu-ci.yml (#13597)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 52, "deletions": 259}}, {"node": {"oid": "bbedf2c4c5abb2d52859313301b38726598a137b", "messageHeadline": "Improve cache locality and perf of DeepGru on CPU (#13582)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 356, "deletions": 69}}, {"node": {"oid": "e0361e62565eff310bfcaf4298c54387a6a06f8a", "messageHeadline": "Change protobuf pin in training requirements (#13596)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d5d69246885befb4c52ee9deef8946121834a00b", "messageHeadline": "rocblas alt impl during backward pass only (#13352)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 138, "deletions": 6}}, {"node": {"oid": "d10d66cc84be67c2d7b4b2bd323e768560c4a390", "messageHeadline": "Cjian/c4244 round 1a (#13483)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 155, "deletions": 153}}, {"node": {"oid": "3482180ec293199444dbf388264ceeba42628e53", "messageHeadline": "DML EP add a registration for Shape and Size (#13442)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 254, "deletions": 44}}, {"node": {"oid": "a9a9c34d986424bff257033601b0f9efe43cd48a", "messageHeadline": "Fix WinML Test Case: create LearningModelBinding for every testcase (\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "281f199754d5ca871faa487e59a49afd247cb5cc", "messageHeadline": "[EP-Perf-Dashboard] Reduce script excessive output (#13562)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 546, "deletions": 407}}, {"node": {"oid": "123e1eac019b1c25da6819254f962cc08ddb060d", "messageHeadline": "Remove torch and valgrind from inference pipelines (#13568)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 13}}, {"node": {"oid": "215732f74b341baa6afa2391d756ef44806cd1f2", "messageHeadline": "Ignore saved runtime optimizations when updating ORT format model <v5\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 174, "deletions": 93}}, {"node": {"oid": "b383312f4cb5815e6551fdbcd92201fe8129136a", "messageHeadline": "[tvm] Add support for int8 models, update TVM revision (#13519)", "author": {"name": "Peter Salas", "email": "peter.salas@gmail.com", "user": {"login": "petersalas"}}, "additions": 29, "deletions": 20}}, {"node": {"oid": "9e65f3bfdbb146c73410489f68a5ebf9a78aecbe", "messageHeadline": "Replace deprecated Python dependency sklearn with scikit-learn. (#13585)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 8}}, {"node": {"oid": "efcbdac58e542273c7cd42f5e5bcd01c0218f0e3", "messageHeadline": "Remove the cmake option: onnxruntime_DEV_MODE (#13573)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 101, "deletions": 79}}, {"node": {"oid": "6201593f24087efac69819d265515cadd20bb655", "messageHeadline": "Remove the dependency on CentOS EPEL (#13567)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 102}}, {"node": {"oid": "23da468154a35134e2c85c3417a92766f2c1b76c", "messageHeadline": "Upgrade cmake version to 3.24 (#13569)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 39, "deletions": 41}}, {"node": {"oid": "8b9065a3967dbf2d8205a18f881d488febc3427a", "messageHeadline": "Add getter/setter of C# OrtEnv log level (#13402)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 70, "deletions": 5}}, {"node": {"oid": "0296bc74c1e92a80a7f465b68069f1ec48b01737", "messageHeadline": "oneDNN ep bf16 enabling (#13484)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 2654, "deletions": 86}}, {"node": {"oid": "4401f50c5ea109291b5b134491dadf8eb4b144cc", "messageHeadline": "Change GSL download to use HTTPS URL. (#13563)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "ab9ac2acc48b28b153dd614eeeefe1e32217322a", "messageHeadline": "Add guidelines for ORTModule (#13553)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 250, "deletions": 18}}, {"node": {"oid": "433f262dd551e79f6b3af6d777b5c94eb907622a", "messageHeadline": "Disable some tests for 32-bit Windows (#13551)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "1977b7ed6a68cfb50721f81445424cd9bf0eb3f9", "messageHeadline": "Fix pythonop training_mode in evaluation mode (#13514)", "author": {"name": "zhijiang", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 75, "deletions": 3}}, {"node": {"oid": "df796bbb6236100fa2de3ef303cf9ea12bc91d64", "messageHeadline": "cast logits to half when T=MLFloat16 (#13454)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 19, "deletions": 5}}, {"node": {"oid": "b4a1ae835046f0c5a5b8fad4fc86b9a4cceee272", "messageHeadline": "Use narrow instead of gsl::narrow. (#13555)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "2de883c5923a1c3cd4d72a5f69a708595bd81c2a", "messageHeadline": "Update CK and fix performance issue on dev machine (#13531)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 49, "deletions": 63}}, {"node": {"oid": "7c3a23c18600d6e2d0755b95e86e3a86bfad36ac", "messageHeadline": "extend some timeout value (#13552)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "a3e7da60e7a12a3fd0f29bc180840f0bcd7f6a85", "messageHeadline": "Trade subgraph recompute for memory (#12852)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1485, "deletions": 5}}, {"node": {"oid": "77be22f3791fde328745661148266be15e030cce", "messageHeadline": "[oneDNN ep] Update from oneDNN v2.7.0 to oneDNN v2.7.1 (#13536)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b1e1b25e04b0ae5f5e0af003933850e6c2122d8c", "messageHeadline": "Delete CUB (#13534)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 48}}, {"node": {"oid": "5914a7e0aef43c67bcc4fc04a519f0313aa3f2b0", "messageHeadline": "Fix an error in the python packaging pipeline (#13538)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b5904c40dd781f13739d471ba088911386f2c01d", "messageHeadline": "Enable ORT in TorchDynamo (#13259)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 837, "deletions": 3}}, {"node": {"oid": "6740528b98bd0ce784d8e8c089829241d1e3e9c3", "messageHeadline": "[ROCm] Fix bug for rocm ep build using MS GSL 4.0.0 (#13525)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c8886c5b4cf11459a27748ad09c4380d5187b682", "messageHeadline": "Revert \"Update CK and fix performance due to lacking -amdgpu-early-in\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 16, "deletions": 11}}, {"node": {"oid": "c557a558160bd6f83c82f05a1bdb667cceb60b85", "messageHeadline": "Fix on-device training ExportModelForInferencing api (#13510)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 20, "deletions": 14}}, {"node": {"oid": "17f0ffd1c82f1bef0977b42515ccbc644597369d", "messageHeadline": "Support More Cases in NoOpElimination (#13460)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 307, "deletions": 31}}, {"node": {"oid": "3d0db47c176e7e8df041bca8a528e7b15b93f7ce", "messageHeadline": "[DML EP] Fix variable shadowing in EinSum (#13520)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "74f905b237dbe1af3e8bdab3574d0c03ecfee6af", "messageHeadline": "DML EP enable the provider in the op tests (#13441)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 1181, "deletions": 56}}, {"node": {"oid": "9d867a07c0c3392c41a33d164c7b8fabf290f43b", "messageHeadline": "Fix regression in CustomOpApi::GetTensorData (#13450)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 18, "deletions": 15}}, {"node": {"oid": "2ecd1d662203d3fb5d77bf2aa97f6b8bcd7d1759", "messageHeadline": "Switch GSL to MS GSL 4.0.0 (#13416)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1127, "deletions": 4485}}, {"node": {"oid": "7fbfbf789fc5851c38e14b7c7b0e141c305162b1", "messageHeadline": "Increase timeout for binary-size-checks-pipeline. (#13498)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "33b8778a46ac29bd18bdb3a9a6feb8f6283210af", "messageHeadline": "Minor improvement for the documentation of kernel explorer (#13490)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 2, "deletions": 6}}, {"node": {"oid": "943e156f4ce7f41d231e4fb0430c91b72047ff22", "messageHeadline": "Allow custom ops to set input memory type (#10879)", "author": {"name": "Fei Hu", "email": "hufei68@gmail.com", "user": {"login": "feihugis"}}, "additions": 138, "deletions": 0}}, {"node": {"oid": "1b494daffa9e1f52d18d3eadea185e3151d6489c", "messageHeadline": "Add yml file for Snpe EP build (#13494)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 310, "deletions": 91}}, {"node": {"oid": "689e524c58e2c3598be9d30b2474433844119593", "messageHeadline": "Move DML packaging pipelines to aiinfra-dml-winbuild machine pool (#1\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 236, "deletions": 407}}, {"node": {"oid": "49e5a11ccde963838333428a6ad2fd939c01ba10", "messageHeadline": "Fix SDL and Prefast Errors (#13465)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "0a524cfe1cfe969b416321af5bc62625f4112365", "messageHeadline": "Fix the input shape of FastGelu (#13488)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "4dd053cc1572ce36f5240cf2492c88fdf7cbe8d9", "messageHeadline": "Update CK and fix performance due to lacking -amdgpu-early-inline-all\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 11, "deletions": 16}}, {"node": {"oid": "8b0669bf63d18cf6a90416d68af29cd306bcbbf7", "messageHeadline": "QuickGelu Fusion (#12417)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 752, "deletions": 26}}, {"node": {"oid": "20c3c35c335d89b5432f157e6ef35aafb3fccee7", "messageHeadline": "[XNNPACK] support building xnnpack EP for IOS (#13461)", "author": {"name": "JiCheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 69, "deletions": 16}}, {"node": {"oid": "07271b6c8a63c6e74574cb1d971801173c34ad11", "messageHeadline": "Update docs/OperatorKernels.md (#13485)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "f9378c5cca407f706fad351bc7892b45d192dc1c", "messageHeadline": "Cjian/c4244 round 2 (#13473)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 104, "deletions": 97}}, {"node": {"oid": "4a20c0d98b84f38974fc709fb1892083e618d023", "messageHeadline": "Delete zlib.cmake (#13467)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 47}}, {"node": {"oid": "67074851a384f0442657ccd6b956e16c898fa4e1", "messageHeadline": "Skip failed models on training ci and openvino ci (#13477)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 26, "deletions": 2}}, {"node": {"oid": "35659d902117add9c97c56b606e496e164e7935e", "messageHeadline": "Increase the timeout value for linux-gpu-tensorrt-ci-pipeline.yml (#1\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ab71c4bbc00ff85818951f98460978252188e2ac", "messageHeadline": "Document generation CI is broken (#13308)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 92, "deletions": 49}}, {"node": {"oid": "0b29f64dba628db012ea71a1ae319dc7fa05d4d8", "messageHeadline": "[DML EP] Enable all datatypes for Abs and Sign (#13470)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "0e2087acfffa9c12a4db3350742eaad82120c29c", "messageHeadline": "Add extension method to compensate for Contains() absence (#13466)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 26, "deletions": 3}}, {"node": {"oid": "a46c599a404bea970fa7948e23d7bc91bbe6dc9e", "messageHeadline": "Training API to export the eval model to an inference model (#13345)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 463, "deletions": 129}}, {"node": {"oid": "8827c4bdbc222dc149b494645e58b4d6164a7070", "messageHeadline": "First round of fixes. (#13452)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 150, "deletions": 150}}, {"node": {"oid": "601b74b904c0310a017672f3160c390fb633a582", "messageHeadline": "Add '$schema' entry to cgmanifest.json files. (#13444)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "7d5833229877ad09265a75983d005d0f39457209", "messageHeadline": "Update tsaoptions.json: update the email alias (#13448)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "805ec459a0c306ca0d89c86e66d676a787248af8", "messageHeadline": "Fix a PoliCheck finding in _hierarchical_ortmodule.py(#13462)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "490e4ddea527a5a6d2c1c7d705bb9cc1d535aaba", "messageHeadline": "[DML EP] Don't fuse a capability outside the compile call (#13468)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 2, "deletions": 13}}, {"node": {"oid": "1c8a22ec68513dd272053fce18fc22cb44a6ece5", "messageHeadline": "Improve logging and default affinity mask generation (#13338)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 253, "deletions": 97}}, {"node": {"oid": "136e15bfaf1c488e5fd0d3e51dda5772df79a3d8", "messageHeadline": "revert cmake external file (#13459)", "author": {"name": "Rui Ren", "email": "ruiren1225@gmail.com", "user": {"login": "rui-ren"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "8770201e96f5875ab075f782624b3c5d275eb906", "messageHeadline": "[EP-Perf-Dashboard] Decouple docker image name from branch name (#13449)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "48b2ec944c89c2d3ec22ee5cfc2285851e02bb04", "messageHeadline": "Fix warnings preventing Onnx build (#13447)", "author": {"name": "Juan Villamizar", "email": "100237675+jpvillam-amd@users.noreply.github.com", "user": {"login": "jpvillam-amd"}}, "additions": 22, "deletions": 23}}, {"node": {"oid": "8fbdc6cc46508420c29f4e4004f0fa19668c8ed0", "messageHeadline": "Add a script for quick profile analysis (#13423)", "author": {"name": "Abhishek Udupa", "email": "abudup@microsoft.com", "user": {"login": "abudup"}}, "additions": 275, "deletions": 0}}, {"node": {"oid": "a0cc289be6408b0180492643f9ec7a81021e89bd", "messageHeadline": "Update SkipLayerNorm fusion rules (#13350)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "ac48bdec891942f7467fe66699e97fd3cdf6f2a4", "messageHeadline": "DML EP add einsum MatMul NHCW ops (#13440)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 141, "deletions": 29}}, {"node": {"oid": "d5e8d59243e98c0da42c8abf5fd268752dcfa5b8", "messageHeadline": "DML EP register all data types for Where operator (#13443)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "70b73afd363fe866060f29ed029b4fb25b335498", "messageHeadline": "[ADD] fuse Matmul + fastgalu -> gemmfastgelu (#11699)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 115, "deletions": 1}}, {"node": {"oid": "cf8bf0c1415e8f454776435549c50ef4b2ec1bc2", "messageHeadline": "add on device training to the packaging pipelines (#13446)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "7aafd862293d53eeceef0e0f57de6b844b1d0586", "messageHeadline": "Update Attention operator to support separated Q/K/V inputs (#13410)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1639, "deletions": 943}}, {"node": {"oid": "a396a91c9ac177a75a2f2b9acfcd629a130227c0", "messageHeadline": "Move build machines with Nvidia M60 GPUs to Nvidia T4 (#13170)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 168, "deletions": 163}}, {"node": {"oid": "0201cd75e103599856f1ae01290797e61661a1f0", "messageHeadline": "Document generation for operator kernels, enable internal overload of\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 13, "deletions": 3}}, {"node": {"oid": "d80212d42c6e60b0a2f0a30f25726303dfb48ced", "messageHeadline": "Add script for question answering (SQuAD) accuracy evaluation of BERT\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 270, "deletions": 0}}, {"node": {"oid": "d82036dbbd0136348d5360b0c3f8231154dcf627", "messageHeadline": "Add Pre- and Post-tunning API to allow pre- and postprocessing of par\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 58, "deletions": 18}}, {"node": {"oid": "b6a3562ffb6d2120cc018b0a15c59ae78336cb35", "messageHeadline": "[ORTModule] Add Env Variable to Control Disabling Custom AutoGrad Fun\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 9, "deletions": 4}}, {"node": {"oid": "ea1bdb162fc8100df59aceca26d65b111f5de758", "messageHeadline": "[NNAPI] Refactor `Resize` as layout insensitive (#13412)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 57, "deletions": 22}}, {"node": {"oid": "93f7a97a6d398853a2d3c3e80dd1206024e6cbb7", "messageHeadline": "Exculde hipify option from policheck (#13431)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "28f470c26c2c2f47ee884230914ec582fbccf332", "messageHeadline": "[ROCm] Use SkipLayerNorm original implementation in kernel explorer (\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 121, "deletions": 76}}, {"node": {"oid": "2748f38362f3aa77fe18de7521b52e7d0fd78224", "messageHeadline": "Drop hip_add_library (#13406)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 58, "deletions": 35}}, {"node": {"oid": "e160688a9b215f733b25a4a41679ee04c2b8c687", "messageHeadline": "Skip some failed models winml and training workflows on Windows CPU  \u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 28, "deletions": 5}}, {"node": {"oid": "24818cfd73a744fdbfe70da9b3e02d06febcaced", "messageHeadline": "[DML EP] Attention Kernel (#13371)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 487, "deletions": 5}}, {"node": {"oid": "1885460776b671c4f2b8b9794fc10fd74611e440", "messageHeadline": "skip some models failed in dynamic shape infer (#13400)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "143725604e05062f653c0a5d5bd20a00284b5ebe", "messageHeadline": "Skip some models failed in Windows CPU C# tests (#13395)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 15, "deletions": 0}}, {"node": {"oid": "397edf99186f39b3512323969878dde53bbb7af8", "messageHeadline": "Bumping up version number to 1.14.0 on main branch (#13401)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 130, "deletions": 108}}, {"node": {"oid": "928c9889a36ab13a00737cd439726140ddbab3a6", "messageHeadline": "A few fixes for generative model ops  (#13363)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 49, "deletions": 21}}, {"node": {"oid": "62cc927f05580a8875af1fc93363d4b6d34bc2c6", "messageHeadline": "[ORT+DML] Validate DML EP header files in ORT+DML NuGet pacakge (#13359)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 56, "deletions": 15}}, {"node": {"oid": "a8701c2a592ee017e21a984df0269bfd268e15ae", "messageHeadline": "Test TunableOp GEMM and MatMul (#13378)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 266, "deletions": 69}}, {"node": {"oid": "928c9fc3488fa957581155860f001c04dd6fadca", "messageHeadline": "Hipify during build instead of before cmake config (#13333)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 253, "deletions": 321}}, {"node": {"oid": "bb16ee712ece6871f87092bdbf43daf6b4d24a17", "messageHeadline": "skip 2 models in C# test (#13384)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "7a3486c3ee6ea365b686427440d1d8030642ef4d", "messageHeadline": "enable arm32/arm64 target for .net apps built against OnnxRuntime.ML.\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "bed169192de4a69f0c0d39d83f18bb4c772f82f9", "messageHeadline": "Windows build fix for on device training training. (#13354)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ac5948cb489abb2c32d949a37cb25f2ab0179b98", "messageHeadline": "Fix bug for percentile calibration module. (#13376)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 18, "deletions": 3}}, {"node": {"oid": "fc12abf6b1e66a3a09622210ef5d5c4574e070fe", "messageHeadline": "Enable/Disbale tunable GEMM by using tunable switch in provider optio\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 137, "deletions": 8}}, {"node": {"oid": "4b2b5888953aef8df0703678973de333cb8b5a68", "messageHeadline": "[ROCm] Fix azcopy issue on ROCm ci pipeline (#13365)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "24b25df64191a86740b303727bf8865e06a40595", "messageHeadline": "Add verbose level log for TunableOp (#13369)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 15, "deletions": 0}}, {"node": {"oid": "665fb346ab29f5a863b90435f0f1418c469b904d", "messageHeadline": "[ROCm] set parallel=16 when build on ROCm CI (#13368)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "67150baa8d50e74f3a8a7b8e679a8d31eae4c0ed", "messageHeadline": "[ORTModule] ATen Support for aten::upsample_nearest (#13364)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 130, "deletions": 32}}, {"node": {"oid": "b6b3f4163619e57b6a8059fccdce89bd51bbc101", "messageHeadline": "Fixes of Hierarchical ORTModule and ORTModule PythonOp (#13347)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 338, "deletions": 104}}, {"node": {"oid": "418304743dada868b8cc751444a8af630aa413f4", "messageHeadline": "[EP-Perf-Dashboard] Update table schemas (#13327)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 493, "deletions": 126}}, {"node": {"oid": "86c5c07ea419b935541d0abd2a16bf204e19c6ed", "messageHeadline": "TRT EP race condition fix during ep compile time (#13356)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 64, "deletions": 62}}, {"node": {"oid": "565da712750ea7fd2c56ee71bae84fa557c13aab", "messageHeadline": "Make 'env' argument to Session const (#13362)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 68, "deletions": 42}}, {"node": {"oid": "9efa8e20bbc442023444140ddd895d394d863997", "messageHeadline": "Add Symbolic Shape and Type Infer for aten::group_norm (#13348)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 21, "deletions": 0}}, {"node": {"oid": "2fa18ea77e919ff771fb0d2a1f458058067c944d", "messageHeadline": "[React Native CI] Record more info to debug E2E test (#13329)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 155, "deletions": 56}}, {"node": {"oid": "9189ebb41591fccfd8c4c90af2db36f9b53b4ffe", "messageHeadline": "Optimize slicing when possible by copying bigger blocks at once (#13261)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 155, "deletions": 61}}, {"node": {"oid": "f5e3165cc3e1b129f4c81f7441126c127b2d8ddc", "messageHeadline": "Fix move Base::operator= (#13355)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f96f2225262ed9aaa17604aeb3185b98c5dc71d2", "messageHeadline": "Change CPU EP behavior with auto_pad when ConvTranspose output shape \u2026", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "15673b4537af36d4b15ea4f856d1ff721b389d25", "messageHeadline": "Revert \"Fix shape-related issues in FuseConv (#12410)\" (#13353)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 61, "deletions": 340}}, {"node": {"oid": "e3982416d3bb0e31de78ae320a1a8c5b109ba57f", "messageHeadline": "Fix Bug where zero point isn't correct under entropy calibration  (#1\u2026", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "61ee5585b2fe62f9eb19e9b02c96f98d8fbe6862", "messageHeadline": "update the nightly build to use the latest ptca image. (#13309)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "68eff69ab193fa8201ce131168a6cfa78cc8eeda", "messageHeadline": "Add Utils for federated learning scenarios (#13014)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 85, "deletions": 0}}, {"node": {"oid": "b4853a978a7a1caa9d89bc33d547dffc13a7ed26", "messageHeadline": "[ROCm] add rocm python package pipeline with --use_rocm_profiling (#1\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 56, "deletions": 10}}, {"node": {"oid": "c4d3c7003f99d386a44a37646a3fde779253fe8c", "messageHeadline": "Refactor provider test utils (#13272)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 195, "deletions": 144}}, {"node": {"oid": "4a63cd02904a30c661b8d0ffdf9f2a3b943b75d2", "messageHeadline": "Improve thread pool creation failure handling. (#13313)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 141, "deletions": 63}}, {"node": {"oid": "1ab11a111ce0717bfbfaca964d04a017cb9b1752", "messageHeadline": "ppc64le: mlas: fix both MaximumFloat and MinimumFloat to return NAN (\u2026", "author": {"name": "Maxiwell S. Garcia", "email": "maxiwell@linux.ibm.com", "user": {"login": "maxiwell"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "4fe6b236995a8ee4cc05e916950cd6fcc97e60c8", "messageHeadline": "Fix typo OpTypesToExcludeOutputQuantizatioin (#13096)", "author": {"name": "fxmarty", "email": "9808326+fxmarty@users.noreply.github.com", "user": {"login": "fxmarty"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "c4a52820a53b493a815cc9002453fb2d7ee1ff31", "messageHeadline": "bug fix for symbolic shape infer (#13067)", "author": {"name": "donglinb", "email": "65906596+donglinb@users.noreply.github.com", "user": {"login": "donglinb"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "65c67764aee7659847ddcf0138121002bc069ad0", "messageHeadline": "remove line \"ADD model ${WORKSPACE_DIR}/model\" in the amdgpu Dockerfi\u2026", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "a561fde126211aa22255455bbffc7cec8fd2b38c", "messageHeadline": "MIGraphX Execution Provider: Stream Synchronization (#12899)", "author": {"name": "Ted Themistokleous", "email": "107195283+TedThemistokleous@users.noreply.github.com", "user": {"login": "TedThemistokleous"}}, "additions": 55, "deletions": 5}}, {"node": {"oid": "3c08f24efc6a1fc4c4c32a6e3f6cf0777273d89f", "messageHeadline": "uset SearchOption for dotframework (#13321)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "790e3639093d0393a917b21a1955b40e0c90fbb3", "messageHeadline": "Reland: Change ROCm to use tunable GEMM (#13231)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 253, "deletions": 26}}, {"node": {"oid": "074e009e69b8e640e6f8947d67f2206f7a291e70", "messageHeadline": "fix jetson build break caused by https://github.com/microsoft/onnxrun\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "dc324b1d9084349c2d35941c58c0e463544769b4", "messageHeadline": "[LazyTensor] Make LORT Build Again with Latest PyTorch (#13303)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 16, "deletions": 24}}, {"node": {"oid": "db32eacda1050d3fdadcd043ef5b6d16f459f168", "messageHeadline": "make the UNSIGNEDPD_CHECK for Windows only (#13260)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "807b2f4dd520cbe6c5cf22c928876745f5b9495d", "messageHeadline": "[ORTModule] Use Env Variable to Set Provider Option cudnn_conv_algo_s\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 28, "deletions": 4}}, {"node": {"oid": "6fb70a82dfc5af0a33546b131d936eb8bb0afb73", "messageHeadline": "[ORTModule] Update Supported DeepSpeed Version for FP16_Optimizer (#1\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 17, "deletions": 4}}, {"node": {"oid": "afb5f767705b40258ce5a0d0af33823a53440e3f", "messageHeadline": "[ORTModule] ATen Support for torch.nn.GroupNorm (#13293)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 118, "deletions": 13}}, {"node": {"oid": "79ac0231a9c5733f05d12ffdf235b23d66310010", "messageHeadline": "Fix scalar sharing bug (#13299)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 235, "deletions": 121}}, {"node": {"oid": "f0fbff6dd48a73f445bae1a9e52cbab5c32f51a8", "messageHeadline": "Adjust docs to comply with Doxygen requirements (#13302)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 14, "deletions": 7}}, {"node": {"oid": "6895918b1c43f6fff1e701646f64bc3a2bd0dbe8", "messageHeadline": "[ROCm] Revert CI pipeline to ROCm5.2.3 (#13297)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "94224387828e9d2d271c01c52cde3530d5c59139", "messageHeadline": "Objective-C static analysis - use different llvm path to try to find \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cbe4eb65b37453aadd5ebd1f94501bfa7f4b1f29", "messageHeadline": "Add backwards compatibility for all versions of ORT format model in f\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 138, "deletions": 45}}, {"node": {"oid": "67bde18d0ddb69c04167d0143d8480b3161bf46a", "messageHeadline": "Update Win_GPU_CI trigger (#13290)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 14, "deletions": 3}}, {"node": {"oid": "5b0d28b5b59b428c9d21937f17b4143c4f968139", "messageHeadline": "Check for null input (#13286)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "a2658f0784aa25b0cebd60a86e02220ac64a7b55", "messageHeadline": "[ORTModule] Fix Graph Builder for Eval Mode (#13255)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 137, "deletions": 98}}, {"node": {"oid": "0d672e9112f5cba3feabecc6d9cedaf6f42c810f", "messageHeadline": "Enable C# test load models with more complex directories. (#13251)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 16, "deletions": 11}}, {"node": {"oid": "8a3407d54febfefa7d510d6ed2d559e90afd3e29", "messageHeadline": "update file name in the comment (#13275)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "1e55949a70e0bc801f7b3d4872eef87a91af678c", "messageHeadline": "Fix unsound hipify in ROCm EP (#13269)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 31, "deletions": 29}}, {"node": {"oid": "b2353fa737e3aa02a7fe99f8c52112443abf7c09", "messageHeadline": "[ROCm] Add ROCm5.3 to python package pipeline (#13249)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 18, "deletions": 225}}, {"node": {"oid": "fb86edb19f09e3d00bbbf2d43d9a101e064e2a00", "messageHeadline": "Update publish-c-apidocs.yml to use main instead of master (#13281)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "93e0a151177ad8222c2c95f814342bfa27f0a64d", "messageHeadline": "implement cos gradient as a function op (#13227)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 15, "deletions": 0}}, {"node": {"oid": "05acd20a885b238ab6a00c3ce98af922ab00bc58", "messageHeadline": "convert singrad to function op and remove cpu kernel (#13263)", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 4, "deletions": 60}}, {"node": {"oid": "cd2e8b306ca91eaedd50e034995c0e9055cbb079", "messageHeadline": "Replace or remove some characters to meet gtest name convention (#13266)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "febd5facce1a2584bc31cbb306a9e81acbb46608", "messageHeadline": "Change head_size parameter dependent on qkv_hidden_size (#12933)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 194, "deletions": 98}}, {"node": {"oid": "b9e23bd086af3607b650f7e0ca55c9d9ce723d87", "messageHeadline": "[ORTModule] Fix Custom Op Registry for Torch 1.13+ (#13250)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 38, "deletions": 16}}, {"node": {"oid": "6b499db7e13e62f19d49b5b521a1dc497d629b63", "messageHeadline": "increase ios pipeline timeout limit (#13268)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ea128cdb185b20fa5cfc989373ce3ce012e91f66", "messageHeadline": "skip windows GPU check if changes only in doc (#13248)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 121, "deletions": 8}}, {"node": {"oid": "4d25b9c8f050ee1dbc995b8948ab670d91bb7962", "messageHeadline": "[ROCm] Update ROCm and MIGraphX CI pipeline to ROCm5.3 (#13257)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 26, "deletions": 24}}, {"node": {"oid": "2cf5d04e3d1439141aa88a9e58e35dcc714ce80f", "messageHeadline": "Fix clang-tidy(cppcoreguidelines-pro-bounds-array-to-pointer-decay) (\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 16, "deletions": 17}}, {"node": {"oid": "00146b25419041bb5e6a232469ff9164936f78c0", "messageHeadline": "Add onnxruntime_BUILD_UNIT_TESTS=OFF definition to iOS package build \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "25c0a66934375dc7ef8a7d7c8b5e13d716ba6057", "messageHeadline": "Natvis adjustments to make debugging bearable (#13237)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 37, "deletions": 13}}, {"node": {"oid": "06686002551d1b7242ae10386a05b2fc1833127e", "messageHeadline": "Share scalar constant initializer (#12878)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 752, "deletions": 108}}, {"node": {"oid": "e01a8519e0a2a3550f0e0edd609bb5058912b6a9", "messageHeadline": "[DML EP] Re-architect | Partitioning as Transformer (#13131)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 1423, "deletions": 1142}}, {"node": {"oid": "38906625a3bc874a55cc349ccf6d6efdfdcce82f", "messageHeadline": "fix some typo in docs (#13212)", "author": {"name": "garanews", "email": "puntogtg@tiscali.it", "user": {"login": "garanews"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "d411bd277e56ff35ec5211672340c24054d2233a", "messageHeadline": "Increase iOS packaging pipeline timeout. (#13233)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bb1c1332450bf9ff4d66a5e2c7d8ccab83fd89ad", "messageHeadline": "[MicroGraph] Address ROCM warning and build failure (#13234)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "6662ece4a14aa47ec37a0290111d866da3168333", "messageHeadline": "increase timeout to 5 hours (#13226)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "04ba8a7e6e917971dd7f0e870b82b4827b16d079", "messageHeadline": "Introduce Training C++ Apis (#12994)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 302, "deletions": 141}}, {"node": {"oid": "51ac6617f5be599eba2ac6e3d995220c1762d109", "messageHeadline": "Fix warnings and enable dev mode for ROCm CI (#13223)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 52, "deletions": 105}}, {"node": {"oid": "5dae0c477d6c70d91952fb2c11fad37301cde7ae", "messageHeadline": "Deprecate CustomApi and refactor public API for better safety and con\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2682, "deletions": 1898}}, {"node": {"oid": "87f55505b3ff4ab9ca1e96d11cbdf147352dd180", "messageHeadline": "[ONNX] Support huggingface BART to ONNX (#12779)", "author": {"name": "Ti-Tai Wang", "email": "titaiwang@microsoft.com", "user": {"login": "titaiwangms"}}, "additions": 1107, "deletions": 0}}, {"node": {"oid": "814e5cfa4c81b059b9865edcbc33510bdf1b1ebe", "messageHeadline": "[rn] Support UINT8 type for onnxruntime-react-native on iOS  (#13210)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 24, "deletions": 2}}, {"node": {"oid": "b09dd11ecee1be83fd978da14f53db0a1255d51f", "messageHeadline": "BFP schemas: Change block dimension type to Int (#13169)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 569, "deletions": 236}}, {"node": {"oid": "cf075fcbad4cef486110f0bae0b6c15f7e01dd55", "messageHeadline": "Handle edge case in CumSum causing overflow (#13174)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 112, "deletions": 71}}, {"node": {"oid": "4e37464cc5901360ade0f569593479f5a8e4acc3", "messageHeadline": "Add build configuration to binary size checks pipeline. (#13208)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 179, "deletions": 75}}, {"node": {"oid": "c7522e547a1d1e37c53de130be8c0bad4ca6b030", "messageHeadline": "Fixed a minor typo (#13194)", "author": {"name": "Tony Xia", "email": "xia_tony@hotmail.com", "user": {"login": "tony-xia"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dca941795e34720529944234b8b168e29036ada6", "messageHeadline": "Fix prefast bugs:\t1944959  1997925  1997926  1997927 1997928 (#13203)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 30, "deletions": 30}}, {"node": {"oid": "72076b1eb27490a8ac64ea8c1743de7449addaf1", "messageHeadline": "Update ROCm CI to use HIP LANGUAGE (#13214)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 130, "deletions": 66}}, {"node": {"oid": "4fc8f7139ac13ee83d6806111dfd608981ee2f84", "messageHeadline": "Bug Fix - C# API order incompatibile with C API (#13191)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "595a0c8658675a16fb1bdcaa9251b0dcb12d53c1", "messageHeadline": "Disable clang-tidy CI (#13207)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "b6c04f48c131d54a8e0fbe27e66ca7a9f6383737", "messageHeadline": "Fix reshape fusion (#13150)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 11, "deletions": 14}}, {"node": {"oid": "2d50d4be24aca5345cb3ddc6736c619b66a6c450", "messageHeadline": "Update TSA path to new ADO project (#12902)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "c780c4a2b97ef597b10e3796ea22eb11277c4e7a", "messageHeadline": "Fix two prefast warnings (#13211)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "962fee5fe56fa2bc7c16c547da1ff700fc2e1658", "messageHeadline": "Fix typo enviroment => environment (#13195)", "author": {"name": "Tony Xia", "email": "xia_tony@hotmail.com", "user": {"login": "tony-xia"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "9cf98dacb37fbb63c4427fd31f3a1ac05d47785b", "messageHeadline": "Enable command list reuse for Xbox (#13173)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 5, "deletions": 13}}, {"node": {"oid": "81a4efee6caca4826fa531176f3229fdb258f61a", "messageHeadline": "Prefast Fixes (#12952)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 17, "deletions": 17}}, {"node": {"oid": "82786baed1f0087e7adb3d4fc19bf142cfb675bd", "messageHeadline": "[js/web] add 'xnnpack' to EP list (#12723)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 67, "deletions": 15}}, {"node": {"oid": "1342baf1c7fc63527b40a730abb65a1d2ffe467a", "messageHeadline": "refine QuantConfig (#13155)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 79, "deletions": 215}}, {"node": {"oid": "0cf17b1921ce882cb8e8e058cc39d38b2eee0469", "messageHeadline": "Add linux debug training package to nightly pipeline (#13192)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 19, "deletions": 15}}, {"node": {"oid": "68218935b9d53feb8caf394362b41ebd8fb5975b", "messageHeadline": "Fix syntax error in labeler.yml (#13193)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a86b8329d9689ca30f0cb9a3a1768195406ac5bb", "messageHeadline": "Update unsupported ORT format version error message to link to doc on\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "054464dce2f0763a6d4027b70ff5c5f4abb3d7c8", "messageHeadline": "fix XNNPACK on WebAssembly SIMD (#13161)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 15, "deletions": 20}}, {"node": {"oid": "0bf0991fa25f71b45c24d1078627d8a8a93cbeb1", "messageHeadline": "Update labeler.yml (#13186)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "56387c3c31699bf71207090649ad9a3c478a3b4d", "messageHeadline": "Fix SDL Unmatched Annotation Errors (#13162)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "aae35f27598a62670399d6d8802d6841bb3afdc6", "messageHeadline": "Binary size reduction in KernelTypeStrResolver and GraphPartitioner (\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 96, "deletions": 37}}, {"node": {"oid": "b76a65c784c95cd620f9295a0ea0ad88b4e4e1b2", "messageHeadline": "Upgrade the oneDNN ep to use oneDNNv2.7 (#13175)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "eef0a98caedae8f13d08c1efb40a48a1cb56a0d9", "messageHeadline": "Update and fix labeling rules (#13129)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 30, "deletions": 53}}, {"node": {"oid": "c20abcab87360eeea5de72fb951d5466adfacc99", "messageHeadline": "User/brianma/eo (#13152)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 55, "deletions": 55}}, {"node": {"oid": "402e1995f0180bf24941747d90eab387c9c24998", "messageHeadline": "Create clang-tidy CI (#12653)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 75, "deletions": 27}}, {"node": {"oid": "5f1bc8ff5651004134daa9b2615d70005c723efc", "messageHeadline": "Add \"--parallel\" to the build flags of WASM pipeline (#13179)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a862b0cad1e7d5792f18a69038b334a02de71271", "messageHeadline": "increase ios_CI_coreml stage timeout limit (#13157)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dd2aec170d5e44959a8992c2d0ecee0c25abcc4d", "messageHeadline": "Update Coding_Conventions_and_Standards.md (#7705)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 112, "deletions": 0}}, {"node": {"oid": "f3180f3ac86a6bf287091db1dadd44649e1f81cc", "messageHeadline": "[DML EP] Enable graph inside DML Graph (#13073)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 555, "deletions": 500}}, {"node": {"oid": "c93cb8f9493aa81ed64c16da11df8bcae7fc0194", "messageHeadline": "Revert \"Enable ROCm to use tunable GEMM\" (#13160)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 16, "deletions": 213}}, {"node": {"oid": "c8781b77f62661ac9765e8d0b6f6adb539fcf4ca", "messageHeadline": "Decouple use_sequence_as_input_ids from has_hidden_states (#13130)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 25, "deletions": 17}}, {"node": {"oid": "32395e2e16a2a571ba600200193ebc1ad23d1995", "messageHeadline": "Add handling for variadic inputs/outputs in a function. (#13140)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 58, "deletions": 10}}, {"node": {"oid": "4d8510611bfc679a714a06fd35aea1dd23f8baef", "messageHeadline": "Update find_optimizer_opset_version_updates_required.py to use the ON\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 46, "deletions": 5}}, {"node": {"oid": "6c63c1c9ee3588bccaace2aa390e91fbef353956", "messageHeadline": "Multiple Gather to Split Fusion (#13095)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 408, "deletions": 0}}, {"node": {"oid": "3157cdb19ac79cab37784cee76bdd0eaeb29bcd5", "messageHeadline": "[ROCm] Fix MIGraphX ciagent user Permissions issues (#13137)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "5182d6610de4808891f63be1aea6811abae1ac17", "messageHeadline": "Upgrade pytorch to 1.12.1 for training pipelines (#13128)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 6, "deletions": 18}}, {"node": {"oid": "c9a86fa27f4af8cca95215973ddeaf075b867029", "messageHeadline": "Openvino GPU Unit/Python Tests fix failure  (#13122)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 133, "deletions": 31}}, {"node": {"oid": "388d3cf84744bea1c6fc23b652c0ca80b199a8d0", "messageHeadline": "[Java] Fix OnnxSequence semantics (#13012)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 131, "deletions": 478}}, {"node": {"oid": "55ae71c16099f79584231f839eb982e0a3cf2b14", "messageHeadline": "Reduce Objective-C static analysis build time. (#13149)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "c26bb1bb19f2b927fabb49559e8300edcc7fecfa", "messageHeadline": "Allow fastgelu/skiplayernorm profile by pass args from commandline (#\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 82, "deletions": 29}}, {"node": {"oid": "32c2c4b480c3983de216a92b19ce06b9cf75780c", "messageHeadline": "Change ROCm to use tunable GEMM (#12853)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 213, "deletions": 16}}, {"node": {"oid": "5e4ebbd9d9b190e5b5bc3bff4cdf0b726921070e", "messageHeadline": "[ROCm] add MIGraphX ci pipeline (#11569)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 135, "deletions": 0}}, {"node": {"oid": "19774f92309e3bbe29a6837600c1865d75880936", "messageHeadline": "print test case's skip reason (#13118)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 8, "deletions": 7}}, {"node": {"oid": "f99d00fa386436670a1165d527db127c942013b9", "messageHeadline": "Add rel* branches to upload training packages to final storage (#13124)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9a44a69653f63074d71dc24d69838895deecddf6", "messageHeadline": "Refactor NNAPI EP OpBuilder/OpSupportChecker structure (#13065)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 6044, "deletions": 5537}}, {"node": {"oid": "457a53c92f2228f32618016907a6eb984d2cfc44", "messageHeadline": "Fix static analysis warning by making derived classes final. (#13123)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "e19163167ebab17ea06508e1d940f7e264073593", "messageHeadline": "Update React Native documentation to reflect change to use full ORT  \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 69, "deletions": 23}}, {"node": {"oid": "5c89c37f7fb448a5a36e32cdd6f8a1c7c02cbceb", "messageHeadline": "Consolidate enabled/default kernel def type constraints (#13034)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 84, "deletions": 277}}, {"node": {"oid": "440f31668f1669be4eec328d7fbbe8a0456d4675", "messageHeadline": "Labeler: Test /i regex for case sensitivity (#13115)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "13d1a3c0074ba3501c07166004d168ca77251763", "messageHeadline": "[ROCm] add SkipLayerNorm vectorize Regular case (#12821)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 166, "deletions": 25}}, {"node": {"oid": "43766ee36d80e1521c43502a60076e3a7a030399", "messageHeadline": "Fix OLive build pipeline (#13114)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 9, "deletions": 10}}, {"node": {"oid": "94e34ace15f4c7011590d387c5dcabaa3e446d7f", "messageHeadline": "Bugfix for SimplifiedLayerNormalization (#12975)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 91, "deletions": 30}}, {"node": {"oid": "237ccc01c77491841309928fd572314f500c6ce0", "messageHeadline": "Remove one last nuphar reference (#13111)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b25437ec413cc1fe233c2ca6a6c5faafcd1edbb1", "messageHeadline": "Upgrade  protobuf version (#13100)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "073dbba78415fe941c29824df28a48a6b09918ea", "messageHeadline": "skip the placeholder inputs while adding node inputs as sub-graph inp\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 47, "deletions": 34}}, {"node": {"oid": "c746083344b959d4c042d5fd5f07d1f0ee6c625e", "messageHeadline": "use parameter names to specify argument mapping (#13108)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "e3bdba37a811f1cef37440fb0ccbab14c0fe4948", "messageHeadline": "Mitigate prefast static analysis warnings (#13032)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 32, "deletions": 32}}, {"node": {"oid": "77a066c700d8ee936c19eed1214527a324c75257", "messageHeadline": "Drop nuphar from java API (#13107)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 43}}, {"node": {"oid": "0e98fb4e9ba02d83f9e92f46e8a33bd05102320b", "messageHeadline": "Fix Build Error for CUDA113 Introduced by 6efa9d9 (#13089)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 16, "deletions": 16}}, {"node": {"oid": "b62ba0b5a78258ff186e971ba7ec3f6d1a515ceb", "messageHeadline": "Remove old enable_linux_gpu_tests parameter from template invocation.\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "e9b1bbc6a5dea915071ce866e4c078fb739d08f4", "messageHeadline": "fix Numpy array None judgement bug (#13103)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 9, "deletions": 2}}, {"node": {"oid": "a83a9ed6b01374a5e1e07b46dbf2fb8e3c407483", "messageHeadline": "Remove miscellaneous nuphar configs (#13070)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 12, "deletions": 175}}, {"node": {"oid": "44c14e8cbb2c13ca487b0468605975004c392aa8", "messageHeadline": "Adding test case for conv per channel with QDQ format (#13041)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 13, "deletions": 1}}, {"node": {"oid": "2ae33b361309ad9c1f9eed4bcb16297e3ce7077e", "messageHeadline": "fix CuDNN lib path for Windows (#12974)", "author": {"name": "Dale Phurrough", "email": "dale@hidale.com", "user": {"login": "diablodale"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ce2ea44a56d08231a9bd6c65229ff84ff03bd31a", "messageHeadline": "Try to fix GitHub labeling action (#12999)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "7116825aef4fbe0bc0e9b8ac1529ee497eced191", "messageHeadline": "Add CMAKE_CUDA_ARCHITECTURES list to python packaging pipeline (#13081)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 10}}, {"node": {"oid": "ade0d291749144e1962884a9cfa736d4e1e80ff8", "messageHeadline": "Updated Dockerfile.ubuntu_openvino with OV 2022.2 official release (#\u2026", "author": {"name": "mayavijx", "email": "mayax.vijayan@intel.com", "user": {"login": "mayavijx"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "365a01397dbd1293e0c2773380c57fd271432b72", "messageHeadline": "Bump protobuf from 3.17.0 to 3.18.3 in /tools/ci_build", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b820256f34d94a4f58512dd6fcffdc7ab9a4df8a", "messageHeadline": "Add check that bias and scale sizes match norm_size in LayerNormaliza\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 36, "deletions": 1}}, {"node": {"oid": "19c51376c42014dc57eb9c7296eea79711e2b781", "messageHeadline": "Introduce QDQ transformer fusion tools for ordered quantized ops (#12\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 978, "deletions": 2}}, {"node": {"oid": "6587a85f8f268946c824b42ef04f650a999fe8f3", "messageHeadline": "Bump protobuf from 3.18.1 to 3.18.3 in /tools/ci_build/github/linux/tvm", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c1ff4b468dfc718f2ceb894a661d68c277cff864", "messageHeadline": "Bump protobuf in /tools/ci_build/github/linux/docker/scripts/manylinux", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9abd6e3a302ff454b7909a33dfea5278377b0124", "messageHeadline": "setup.py: use packaging instead of wheel.vendored.packaging (#13083)", "author": {"name": "Chih-Hsuan Yen", "email": "yan12125@gmail.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2cc4e7e5c2d528179b8b63afdc2c5a09db162df8", "messageHeadline": "[Build] Fix broken AMD CI (#13082)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "63c3b21902b4bb77f28e532d7b6b20f50245bf49", "messageHeadline": "Bump protobuf from 3.18.1 to 3.18.3 in /tools/ci_build/github/linux/d\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8e2528bad24fde3f5d3f90047408140be975f9f3", "messageHeadline": "More LayoutNormalization opset 17 changes (#13066)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 419, "deletions": 314}}, {"node": {"oid": "9e21ffb64984d18ce54fc648fdf863eed7e9399d", "messageHeadline": "Add license header to some files. (#13074)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 3}}, {"node": {"oid": "bcc93ab17c3151ff34abbc068946eaf4fb889434", "messageHeadline": "Deprecate ORTTrainer (#13022)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 17, "deletions": 9}}, {"node": {"oid": "6f27659ceb38ddf1f92ac92c070b170658e4d9cc", "messageHeadline": "Fix prefast warnings (#13017)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 26, "deletions": 14}}, {"node": {"oid": "8bb16ab90012942316b15dfe1923365b0baeb4fc", "messageHeadline": "Propagate environment variable to docker image (#13031)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "6efa9d9e106c818020215988046aec86a829baf5", "messageHeadline": "Add more qordered int8 operators for CUDA provider (#12949)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 3742, "deletions": 284}}, {"node": {"oid": "5f611b63a1a2c9ea7a87ccc2c634e920d8ca962a", "messageHeadline": "Make classes IKernelTypeStrResolver and IKernelLookup have protected \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "2ef1f8b93ea7a3717d7e95d51963947364f003f7", "messageHeadline": "[ROCm] add tunable SkipLayerNorm for ROCm EP (#12817)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 105, "deletions": 27}}, {"node": {"oid": "eafd67b8fd9332062a96d0e21083396f252f1eed", "messageHeadline": "Update CUDA version to 11.6 and refactor python packaging pipeline (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 503, "deletions": 486}}, {"node": {"oid": "92237567d3f3464fbc2de50886ea3f15aeeb0a9a", "messageHeadline": "add opset17 node test data (#13062)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "a24b41d92e2aaa51087f52dbf57d2d7f6e7dfd82", "messageHeadline": "Move all TunableOp related falicilities to EP level directory (#12857)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 361, "deletions": 281}}, {"node": {"oid": "8fb3f05cd65426a32a6e8bc30bdb33843e3a2903", "messageHeadline": "Add cgmanifest file in codeowner list (#13042)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "394c249c7cd2634a74f1eef55a108a6f5a1bff54", "messageHeadline": "Add ONNX LayerNormalization(17)  (#12978)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 247, "deletions": 158}}, {"node": {"oid": "952c99304a764328070ac6c07f510deec06dedf4", "messageHeadline": "Add CANN EP (#12416)", "author": {"name": "wangxiyuan", "email": "wangxiyuan@huawei.com", "user": {"login": "wangxiyuan"}}, "additions": 3876, "deletions": 2}}, {"node": {"oid": "078ceab1dba970c1598548c01642e6e2517d9884", "messageHeadline": "Use full ORT package for onnxruntime-react-native. (#13037)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 119, "deletions": 40}}, {"node": {"oid": "c4a7e88fc8c9f28fd58f1a008a9313a3ab7aee8f", "messageHeadline": "QuantizeBFP and DequantizeBFP (#12833)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 1200, "deletions": 1074}}, {"node": {"oid": "057567f39f39bcc77b7ea3744dc0b619e627a4eb", "messageHeadline": "Fix bug in Attention Fusion (#13050)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cccbe90764616749f06633fadaf90aba2eb360c9", "messageHeadline": "Openvino ep 2022.2 v4.2 (#13023)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 452, "deletions": 489}}, {"node": {"oid": "6ea8780886187e0ae32bb7ffee0b77161e1c4d2a", "messageHeadline": "Replace std::exclusive_scan() with for loop because std::exclusive_sc\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "c7a4093db8d78fd9a98e836292170b41ab688f54", "messageHeadline": "Fix prefast static analysis warning by not calling delete explicilty.\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "2f9b358d16885c8330eaa34a932315e3b176082c", "messageHeadline": "Replace the source of TRT version and fix the build  (#13046)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2f9b559391de2076181b23bc97b9094532066e17", "messageHeadline": "Declutter pull_request_template (#13026)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 9, "deletions": 4}}, {"node": {"oid": "851b0ce93696aeb44bbc8e4cc4ea4d1c10adfe29", "messageHeadline": "[js/web][Fix] - updating the C API to catch non-tensor data (#12811)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "8de5535e9cd8cbd2ad9ca55be0c665d1f04ba2e3", "messageHeadline": "Reduce test warning spew due to CPU fallback (#13035)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "051a0a67a5a5ea5009e36329f6a3081022fce851", "messageHeadline": "Cjian/per channels not working (#13038)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6248b6979572fae99f9ac88466a0c26eb0cb9483", "messageHeadline": "Fixes bug which makes quantized_input_names = [] (#13029)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 29, "deletions": 11}}, {"node": {"oid": "240aeadf1a875cebbd940c125df34ea472cca1e8", "messageHeadline": "Update engine hash id generator with model name/model content/metadat\u2026", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 561, "deletions": 1}}, {"node": {"oid": "39e20686a0c4ff526d580098807b1c2018f98d37", "messageHeadline": "[EP Perf Dashboard] Fix incorrect calls to trtexec with fp16 inputs (\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 27, "deletions": 16}}, {"node": {"oid": "3810effe6e57544ca386046d7f56b7e8a281b590", "messageHeadline": "[NNAPI EP] Remove/Refactor shaper inference calculation code (#12618)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 357, "deletions": 888}}, {"node": {"oid": "8356e3b9b03d5f4dc99fc0ced80b65373daa8b08", "messageHeadline": "Add onnx single node test data to tests (#12822)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 95, "deletions": 96}}, {"node": {"oid": "a5d70d8609e6c35a2b75007cdd422beb09752e29", "messageHeadline": "Allow bert_perf_test.py make some noise by log_severity option (#13024)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 17, "deletions": 1}}, {"node": {"oid": "e9d91cac55ca7245cf60e40dfaf917af1632a49d", "messageHeadline": "Fix hipify not running if the pwd is not the root of onnxruntime repo\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "b2b4f703a54ad17ca1e186cd92787912e95bf15d", "messageHeadline": "Move Linux GPU CI pipeline to T4 (#12996)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "bee49dd112f764a48fa4bbabb805d187cf8a6030", "messageHeadline": "Add limit input rank <= 4 in NNAPI EP Sigmoid op support checker (#13\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 81, "deletions": 2}}, {"node": {"oid": "40749124b132a8a114a639548eec76abe6152e57", "messageHeadline": "Fix Deferred Release and Add New Test Framework for CUDA EP-specific \u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 199, "deletions": 26}}, {"node": {"oid": "4113df0e211ae39724d2d76106a34a64df7e2ee5", "messageHeadline": "use constexpr (#12953)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "dd39f0293df3289321b5b178facea628e147f694", "messageHeadline": "fix static analysis: integer_gemm and attention_quantization (#13004)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 13, "deletions": 11}}, {"node": {"oid": "454f77cd94901cb95c92b20c60565408b2be045c", "messageHeadline": "Update kernel matching logic: decouple from op schemas and remove ker\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3282, "deletions": 6178}}, {"node": {"oid": "32878a1e5835529db144f68f663b60c55ad2e74d", "messageHeadline": "Fix log timestamps being off by one hour when DST is in effect. (#11385)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "1245c6397ec7a84bb839f5537679d1bc6354e3d4", "messageHeadline": "Remove usage of torch.onnx symbolic_registry (#13011)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 12, "deletions": 9}}, {"node": {"oid": "189aef2beaecdbdd3a68fbf569d9891c11d9777c", "messageHeadline": "[ADD] add skip layernorm to kernel explorer for ROCm EP (#12816)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 407, "deletions": 127}}, {"node": {"oid": "ffeba98a9d95b1478d39bade838ac7a90aaf5137", "messageHeadline": "Allow gemm profile by pass args from commandline (#12991)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 42, "deletions": 14}}, {"node": {"oid": "f26054decaf301ef8ac04f062c9bbc23c3f837bf", "messageHeadline": "[XNNPACK] Support running in multi-thread with seperate pthreadpool (\u2026", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 89, "deletions": 38}}, {"node": {"oid": "a8b0f57d1a2e84e6570b2f328405753316013220", "messageHeadline": "Fix eager mode pipeline to accommodate recent allocator change. (#13000)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 84, "deletions": 85}}, {"node": {"oid": "0ddf4efbd966d4189eff6b05cd8d4dc3fa52be9a", "messageHeadline": "Make PythonOp report dtype mismatch by name, instead of by using enum\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "77b567df665dd43b07c775dd0500b947114571c3", "messageHeadline": "test qdq loss presence (#12928)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "3cd2d4a7a108c2b6145a9df387fc693255065d6b", "messageHeadline": "Merge pull request #13013 from microsoft/prathikrao/setuptools-versio\u2026", "author": {"name": "Prathik Rao", "email": "prathik.rao@gmail.com", "user": {"login": "prathikr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8ea742b50721e8c8f781d2e531f963344286b0f4", "messageHeadline": "downgrade setuptools", "author": {"name": "Prathik Rao", "email": "prathikrao@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "14eb3cf48516c1161024c382b370e380f205cf25", "messageHeadline": "Ignore settings.json in git (#12988)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "14365b67a0de0acc1e6423477d7363bd5a98f0d7", "messageHeadline": "Fix hipify due to CUDA EP tensorrt_fused_multihead_attention optimiza\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 35, "deletions": 8}}, {"node": {"oid": "e02bea2e3fcf94ab91131526928d89cce3c8aa33", "messageHeadline": "Fix some warnings (#12918)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "4ed5a5b2a80a938d68193bb5a4c99f40a7cee93b", "messageHeadline": "Disable local versions based on environment variable (#12997)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 19, "deletions": 11}}, {"node": {"oid": "b48f71fcfc04490c73397562ecd0b64c0ad45e87", "messageHeadline": "fix bug: quantization shape inference (#12983)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 40, "deletions": 5}}, {"node": {"oid": "1a684152ccefab5b5c33e6bd66a16ef257940ef0", "messageHeadline": "Fix C6011: dereferencing NULL pointer with_data (and external_data) (\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 24, "deletions": 13}}, {"node": {"oid": "12aab3c01d87c51a9237d553b6135f49f555f256", "messageHeadline": "Fix TSA warnings (#12950)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 38, "deletions": 46}}, {"node": {"oid": "268bfe2a5da40f6e24943d960116154f589d7f29", "messageHeadline": "python training api bindings (#12610)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 432, "deletions": 8}}, {"node": {"oid": "2b5b11d373ec7cd5aa0e3ee4b3dbe70a67bc5ab5", "messageHeadline": "[C#][TVM EP] Fix issues related to using TVM EP in C# front-end (#12958)", "author": {"name": "Alexey Gladyshev", "email": "wotpricol@mail.ru", "user": {"login": "KJlaccHoeUM9l"}}, "additions": 89, "deletions": 3}}, {"node": {"oid": "85546255c4d004438ab7dcbf809ed8032f931ab2", "messageHeadline": "make nhwc transformer only apply to CPU ep.  (#11882)", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 39, "deletions": 16}}, {"node": {"oid": "ab45ac311fc83c62be6e0c69da1e631ab28e1afa", "messageHeadline": "Merge pull request #12980 from microsoft/WindowsAI", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 625, "deletions": 91}}, {"node": {"oid": "b935524e22fdce64874df94ecfb549cc0c87fd7f", "messageHeadline": "Revert reverse setup of allocators + create/register allocator in CPU\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 72, "deletions": 76}}, {"node": {"oid": "94d9e9ad6d18c2846868891ee17ad0c4bdea5de9", "messageHeadline": "[Issue labeler] Separate out C# api as separate label (#12951)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "709254949a213d5ed652d026e7e784efc9820c86", "messageHeadline": "DML EP temporarily fall back to CPU for LayerNorm when Bias is not pr\u2026", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "3c427a89464d046cfb19a514191aaa6f04750035", "messageHeadline": "Fix an arithmetic overflow warning (#12961)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "739b5675c8d28d3177ecc0f24cb82373014365f6", "messageHeadline": "remove legacy compile api (#12932)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 3, "deletions": 66}}, {"node": {"oid": "203f63c224cb2760cf4cdb6542d50619dee6a4e8", "messageHeadline": "Publish WinML Nuget package to ORT-Nightly ADO feed (#12904)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 36, "deletions": 5}}, {"node": {"oid": "9f6646f11d17f9a0d0eea771d79ab38f7da59e4b", "messageHeadline": "Merge branch 'master' into WindowsAI", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal@microsoft.com", "user": {"login": "sumitsays"}}, "additions": 169918, "deletions": 4243}}, {"node": {"oid": "363c695dad7f35bb7a9cc2c570021c4768f76df8", "messageHeadline": "Update DML 1.9.0 to 1.9.1 (#12966)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "08af88e3e2be722254b7fda1b545922302ca5634", "messageHeadline": "Assign generate document job to CPU pool. (#12973)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "647e09cc3974db3ca83caa0ed7882700521ee3bc", "messageHeadline": "[FIX] skip layer norm for ROCm EP (#12803)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 34, "deletions": 38}}, {"node": {"oid": "d2aa2109c05f4008b09eaf0de1907072bc055644", "messageHeadline": "Make TunableOp follow stream semantics (#12856)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 30, "deletions": 6}}, {"node": {"oid": "248f72e97278da3c70eaace63c269f65b5b77554", "messageHeadline": "fix VC++ Static Code Analysis warnings (#12940)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "10f9a69707c435ea381f2b80d8fda2efa71302c8", "messageHeadline": "Use CMake EXCLUDE_FROM_ALL for composable kernels to avoid building o\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "d819b56fbad542808c1d45bb32426d263a42dc04", "messageHeadline": "Consume ONNX 1.12.1 to prevent vulnerability issue while loading exte\u2026", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3f456a1847e87089f113c4c1317cb59352493b57", "messageHeadline": "[Update] update rocm5.2.3 (#12942)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 17, "deletions": 31}}, {"node": {"oid": "5099dda42f07b5e096971e9598c2a893bff244b2", "messageHeadline": "Lint updates csharp docs (#12962)", "author": {"name": "Cassie Breviu", "email": "46505951+cassiebreviu@users.noreply.github.com", "user": {"login": "cassiebreviu"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "bc2df1bf959bc34773dc63fb261b02d1398312bf", "messageHeadline": "Remove previously deprecated API (#12935)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 63, "deletions": 186}}, {"node": {"oid": "1ef10291633fb29454bdfc6e39c4cb008fff9112", "messageHeadline": "Skip 2 tests in windows gpu workflow (#12956)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "b8e34fbd916283c1da03cf282c8794911a632d0e", "messageHeadline": "Split topk implementation into per-type translation units to speed up\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 80, "deletions": 31}}, {"node": {"oid": "da07c83948a456e75f87fadd97de6d5437d2ec7c", "messageHeadline": "SoftmaxCrossEntropyLossInternalGrad and Sum Fusion (#12746)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 519, "deletions": 43}}, {"node": {"oid": "568950e28cbd03711b07b0ceb72adce514fdbc2c", "messageHeadline": "Warn on node EP silent fallback from preferred provider (#10831)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 35, "deletions": 6}}, {"node": {"oid": "78bc53f91dbf6bb72d79441231879656c6847734", "messageHeadline": "fix prefast:Warning C26814 in non_max_suppression.cc (#12934)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bb98922cc8583b7f81a6f5ed542715dc3eb3ce49", "messageHeadline": "Delete nuphar docker file (#12944)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 31}}, {"node": {"oid": "95c4fc6877fa1257b4d0d4b0f78bc61b6a0bdba4", "messageHeadline": "[CUDA] Add TensorRT fused attention fp16 v2 kernels (#12814)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 168076, "deletions": 198}}, {"node": {"oid": "1016c33519c65ca41aba7b05a78419c3b8cb560b", "messageHeadline": "Fix prefast warning in upsample.cc. (#12938)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 37, "deletions": 35}}, {"node": {"oid": "626d94aa239cc255c3ed90227e2e43872b6ebc43", "messageHeadline": "Refactor python packaging pipeline and nuget packaging pipeline (#12945)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 792, "deletions": 539}}, {"node": {"oid": "9edc9465f0258fa4cf83d7975905afa9ce1ec436", "messageHeadline": "Fix some prefast warnings (#12936)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "64466c2d6206d25de52222a49f0cda0f0a145787", "messageHeadline": "Remove nuphar provider folder (#12939)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 0, "deletions": 2932}}, {"node": {"oid": "28e27ee7f7735f4f4d4126fa6c18ca1dffdda71d", "messageHeadline": "Changes for AIX compilation to get CPU of running thread. hz is inter\u2026", "author": {"name": "madurais", "email": "madhu011989@gmail.com", "user": {"login": "madurais"}}, "additions": 79, "deletions": 9}}, {"node": {"oid": "31a1403e060ece51ebee8014b6282ad214317b89", "messageHeadline": "Add --output_dir option to convert_onnx_models_to_ort.py. (#12844)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 31, "deletions": 5}}, {"node": {"oid": "a433f22f17e59671ff01acf0d270b7e3476a952a", "messageHeadline": "Softmax interface update (#12469)", "author": {"name": "Joseph Groenenboom", "email": "joseph.groenenboom@amd.com", "user": {"login": "groenenboomj"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "30ebc9e00ad44ae00f0474e92437ba8a4ab22570", "messageHeadline": "Useless Cast removal after converting model from float32 to float16 (\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 66, "deletions": 110}}, {"node": {"oid": "d8636c2be88deb529ac42497c3dc3195e19c442a", "messageHeadline": "Add enable_onnx_tests in windows nuget test step (#12926)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "f78ed1388a21380ebd62bf2bd354138c123b62c7", "messageHeadline": "Fixed build break: inbox version of WindowsAI repo", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal@microsoft.com", "user": {"login": "sumitsays"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "bcdddb47bafabe60c16d319651431441abf07587", "messageHeadline": "Merge remote-tracking branch 'origin/main' into WindowsAI", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal@microsoft.com", "user": {"login": "sumitsays"}}, "additions": 37198, "deletions": 31832}}, {"node": {"oid": "5e978f351cced9388e519b3cf36c26b24865254f", "messageHeadline": "[DML EP] LayerNorm Kernel (#12809)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 86, "deletions": 1}}, {"node": {"oid": "1e34440c370d40085fad5fd1b1002b4a04c5991d", "messageHeadline": "Fix ORT crash when loading BeamSearch model (#12872)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 188, "deletions": 0}}, {"node": {"oid": "022d9e2d0cd94e6bc13a03586980071dbe182897", "messageHeadline": "Get files for XNNPACK wasm build from BUILD.bazel. (#12892)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 57, "deletions": 33}}, {"node": {"oid": "e561a7cf29bc61c1ad2b8276e0096c6d4f4170c7", "messageHeadline": "Adding QuantConfig Class (#12810)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 551, "deletions": 108}}, {"node": {"oid": "8e4eb24648048a0c3c1f3fca3b4a43792286fdba", "messageHeadline": "Update operator kernel table to include DML operators (#12887)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 338, "deletions": 9}}, {"node": {"oid": "05c65a54b33f5ebb0a6a9b1880d9e281dbd9d510", "messageHeadline": "[DML EP] Contrib Op: FusedMatMul (#12898)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 561, "deletions": 115}}, {"node": {"oid": "0b235b27632e96c5f9cf01093d9d4d4ecf2f8447", "messageHeadline": "Disable QOrderedMatMul with bias tests on Windows (#12901)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "b5327595f32dd6f682ba7faa5bfd699aaa55ebe3", "messageHeadline": "Fix [prefast:Warning]: C26814 (#12897)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "5d55b0730e481a967bd2da68f04f0d4efa8d5831", "messageHeadline": "[Java] JNI refactor for OrtJniUtil (#12516)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 697, "deletions": 624}}, {"node": {"oid": "60e4d012e002569371017426aaf0f78462ce616b", "messageHeadline": "Fix unused variable warning from reduced ops build (#12889)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "28f2e57de568c80bbdeeff22b2f9cf71eb68aeb6", "messageHeadline": "Use CUDA callback to release deferred-release buffers (#12883)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 252, "deletions": 149}}, {"node": {"oid": "55c745eefdeea54ff1e527b2634289dde17ddbc5", "messageHeadline": "Add support for ORTModule Torch cpp CUDA extension build within docke\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 12, "deletions": 3}}, {"node": {"oid": "6ebb7b91eba7fb98ee2a234366efd091426a18ea", "messageHeadline": "Re-apply fix for mkl issue for eager mode (#12881)", "author": {"name": "pallavides", "email": "pallavides@users.noreply.github.com", "user": {"login": "pallavides"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "ff52d6a6bfe34de4f327a5815d8aac65a4140cc3", "messageHeadline": "Delete Dockerfile.ubuntu (#12888)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 57}}, {"node": {"oid": "a811c7629fd19b971d6607ad16fb5edf4ba9ba41", "messageHeadline": "Remove \"Build Python Documentation\" from py-packaging-stage.yml  (#12\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 19}}, {"node": {"oid": "b1984278d9b8f6162ccc07c09aa7890581830a86", "messageHeadline": "Enable blank issues (#12885)", "author": {"name": "sophies927", "email": "107952697+sophies927@users.noreply.github.com", "user": {"login": "sophies927"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4765e5c382deb22e0be3b7a05650b6440bd3e32b", "messageHeadline": "Using ORTModule to wrap a evaluation model should not change the mode\u2026", "author": {"name": "guyang3532", "email": "62738430+guyang3532@users.noreply.github.com", "user": {"login": "guyang3532"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "d3b684cd9e36e085ffd4dced826e98cc50ef2bc3", "messageHeadline": "Drop nuphar (#11555)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 85, "deletions": 17009}}, {"node": {"oid": "acc8bdc6c5a4326c75ec5383eb9aa13d45c469ca", "messageHeadline": "Splitting quantize_tensor and quantize_input (#12873)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 213, "deletions": 85}}, {"node": {"oid": "535b0835f267896e6e739fd3614678d4ce90a41f", "messageHeadline": "User/sheilk/dft fixes (#12862)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2933, "deletions": 2875}}, {"node": {"oid": "f63bd0765dd44b9136b06d6ac6c424b3555fec60", "messageHeadline": "New GitHub templates (#12777)", "author": {"name": "sophies927", "email": "107952697+sophies927@users.noreply.github.com", "user": {"login": "sophies927"}}, "additions": 739, "deletions": 54}}, {"node": {"oid": "ad69aac491fc423fc71484a8c633ef444eec3105", "messageHeadline": "Introduce ordered quantization ops for the CUDA EP [1/n] (#12582)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1626, "deletions": 9}}, {"node": {"oid": "69f7cc6494cb88f2b51acee70ff7b1f0c3b0505e", "messageHeadline": "Add pybind support for all memory config options in OrtArenaCfg (#12658)", "author": {"name": "petermcaughan", "email": "peter.mcaughan@gmail.com", "user": {"login": "petermcaughan"}}, "additions": 68, "deletions": 2}}, {"node": {"oid": "8004db4bf18635451870bee8a07b8f80b0bafec1", "messageHeadline": "fix python import sequence warning (#12864)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 143, "deletions": 131}}, {"node": {"oid": "400195a10a1a06caa162e0e024ead17f4f07ce8e", "messageHeadline": "raise an exception when TreeEnsemble request a feature out of boundar\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 57, "deletions": 0}}, {"node": {"oid": "f856be162e321021fd779aeb54de52b6df276ed1", "messageHeadline": "fix xnnpack wasm build (#12845)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "437409c343f5c9cbfab3f4d3e8f82bab992447f4", "messageHeadline": "Add DONT_VECTORIZE flag to cmake (#12169)", "author": {"name": "Jan Tilly", "email": "jantilly@gmail.com", "user": {"login": "jtilly"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "706e03c63df9b5dec4b12c4a6027124bc0496901", "messageHeadline": "Add azp run helper (#12832)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 100, "deletions": 0}}, {"node": {"oid": "c571b993363d1f1095c2921c9ff3f9770c02eb85", "messageHeadline": "Refactor setup_test_data (#12818)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 54, "deletions": 37}}, {"node": {"oid": "726251609a546e8cf05b8023f118bc814778a3a4", "messageHeadline": "increase max memory to 4G for wasm (#12798)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d19955fd89a18dbe82bed4396c7f2923995095fc", "messageHeadline": "fix transformers script issues (#12802)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 63, "deletions": 73}}, {"node": {"oid": "54360c88d21912f63cc1cc9b01df82cd85dab845", "messageHeadline": "Disable two warnings raised by tensorboard on Visual Studio (#12773)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9ad5b95e4f4383df29da9dba6dd4a13e4a506335", "messageHeadline": "Fix math domain error with log10 (#12841)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "8cedafe250ee3b5baaaab854d6cb5b3e05e0254c", "messageHeadline": "[xnnpack] Have `Initializer` in Mobile related EPs in Minimal_build a\u2026", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 250, "deletions": 363}}, {"node": {"oid": "ac4f1bf9603967212333cf9c9e5487145e7c91bc", "messageHeadline": "Update max opset for NNAPI and CoreML.  (#12831)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "9e47eb68e0ec1a7a1476ceaaf69c6c4c58e91090", "messageHeadline": "Remove unused orttraining amd dockerfiles and scripts (#12707)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 0, "deletions": 1804}}, {"node": {"oid": "76d17b0f48f01229c686eee33c22fe70f4a0b9da", "messageHeadline": "Add java API for xnnpack (#12788)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 83, "deletions": 2}}, {"node": {"oid": "295bd26980ff8f399bb544673d995484dc9d14b1", "messageHeadline": "Remove orttraining-distributed CI pipeline (#12738)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 6, "deletions": 5718}}, {"node": {"oid": "27dde0b51f17c65211fb15090f8cba5f2baf8898", "messageHeadline": "Csharp bindings for on-device training APIs (#12404)", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 1371, "deletions": 123}}, {"node": {"oid": "2fe919c3ad0a5f1cb35c126489c65c2e047988f0", "messageHeadline": "Adding Split Fusion (#12732)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 109, "deletions": 14}}, {"node": {"oid": "56bae3b196ad958edc2ae8923e6a554ca30b6cfa", "messageHeadline": "Use InplaceClipGradNorm for offline processing for on-device training\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 262, "deletions": 102}}, {"node": {"oid": "98b2b7f5bb51f8c81e937ad3beedfd946825add4", "messageHeadline": "Update csharp documentation (#12830)", "author": {"name": "Cassie Breviu", "email": "46505951+cassiebreviu@users.noreply.github.com", "user": {"login": "cassiebreviu"}}, "additions": 28, "deletions": 51424}}, {"node": {"oid": "548938fb9771fd51167306f9ac079db7c5c3ea77", "messageHeadline": "Update stale.yml (#12813)", "author": {"name": "sophies927", "email": "107952697+sophies927@users.noreply.github.com", "user": {"login": "sophies927"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "ca5af247654f04992e23d473987c1b3edf860d7f", "messageHeadline": "Update Sdl.ruleset to remove C26812 from the rules (#12695)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "931c8b01479197ba2aa293a609d0e91a221e47bb", "messageHeadline": "Resolve GH issue 12706 (#12815)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 24, "deletions": 0}}, {"node": {"oid": "6fe712b587ae23fb3b5bf3fa2a3a3c52ecc073d9", "messageHeadline": "Create codeql.yml to replace LGTM (#12790)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 57, "deletions": 0}}, {"node": {"oid": "349469c381282432bee76af44ab1acf958917973", "messageHeadline": "Enable way to extract all parameters to and from a contiguous buffer.\u2026", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 296, "deletions": 4}}, {"node": {"oid": "52ce6a90b4b52e0b331e6c0c61e05478589bf108", "messageHeadline": "Props file cleanup (#12782)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 50, "deletions": 11}}, {"node": {"oid": "0125e15281f94f857dfcd8cdc83003d0b0b04dde", "messageHeadline": "Fix include order build failure training build (#12425)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "e3b501125d795c4379f1399afb9651dc0704fbf8", "messageHeadline": "DFT on DirectML (#12710)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 6184, "deletions": 46}}, {"node": {"oid": "5dd9afe75a603792f05531ad082f5a28b0ba260e", "messageHeadline": "python lint (#12825)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "adbc0757ad9f96f8b44d3de526c36917ab6e4431", "messageHeadline": "[UPDATE] update ROCm ci pipeline to ROCm5.2.3 (#12799)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "262a597e2a21ecdc7c3d5bde07c4623045972412", "messageHeadline": "[CUDA] BiasSoftmax and Dropout Fusion (#12667)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1606, "deletions": 694}}, {"node": {"oid": "7e1e0f5c9e9a18a421d0b3369a6d0359ac23998f", "messageHeadline": "fix build error for generation debug code (#12801)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "a48b1155408e516447d0594b0a4bbad90a0348e0", "messageHeadline": "Remove reference to the deprecated variable in `torch.onnx.symbolic_h\u2026", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 28, "deletions": 4}}, {"node": {"oid": "f59a44e24feec730bcb145bd5f377ce07f4cae9e", "messageHeadline": "Build VS 2022 no Abseil adjustment (#12780)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "cc94ba77407c8f499f7a7d7e85d497e6c8ea21ca", "messageHeadline": "Update C# API docs workflow  (#12796)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 6, "deletions": 9}}, {"node": {"oid": "6b148e15e143a2143abf6687ef71dd8578c3e912", "messageHeadline": "Disable QGEMM, s8 A, s8 B (Packed) bench for AMD64 (#12765)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 11, "deletions": 8}}, {"node": {"oid": "a52543ecd8ce1bde9b09e35aa8af5cd79915b11a", "messageHeadline": "Generate windows training package (#12789)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 40, "deletions": 20}}, {"node": {"oid": "82a28cc2c3f9402f3f2947a317d8cfc99a006503", "messageHeadline": "upgrade emsdk to 3.1.19 (#12690)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 28, "deletions": 8}}, {"node": {"oid": "1a402a3f25669e31feb7651a143b93aee5bc979a", "messageHeadline": "replace 'master' branch ref to 'main' for onnx repo (#12678)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 271, "deletions": 272}}, {"node": {"oid": "9aefcc251fec4e3ff85709bebad8a8066e6fd288", "messageHeadline": "fix some prefast warnings (#12730)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "9680ffd8426e630f025925a16d51e5010a270cbf", "messageHeadline": "Fix rocm build caused by #12699 (#12787)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b4f6dad7c9e09c432b04d862729021c360389bf7", "messageHeadline": "increase timeout limit of mac silicon package workflow (#12784)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9907b59a1ea746d6196c8d82ed76b9f0b8201e4e", "messageHeadline": "Change cuda and rocm error checking helpers to return Status (#12699)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 430, "deletions": 630}}, {"node": {"oid": "a0c25e5c2f59d64ab697b5c586d0f0bb9af4a91d", "messageHeadline": "Fix segment fault for alltoall (#12701)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 29, "deletions": 16}}, {"node": {"oid": "19ca2a0089936a2710fc1d651033f1db6eb2a621", "messageHeadline": "[ADD] python package pipeline for ROCm5.2.3  (#12770)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 233, "deletions": 41}}, {"node": {"oid": "d761a7ceb30d12ad0cabb33e04926012410feb8b", "messageHeadline": "Pre-processing of Quantization (#12729)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 312, "deletions": 3}}, {"node": {"oid": "1ce14e752b42b09bcaeb5d818475dc01f2007b3e", "messageHeadline": "Increase timeout for clean-build-docker-image-cache-pipeline. (#12776)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "17ccd6fa02877a1c8d3201344137b1ca105b681d", "messageHeadline": "Fix shape-related issues in FuseConv (#12410)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 340, "deletions": 61}}, {"node": {"oid": "233f8c210e462ae084ec501a20f0a38b73acb7dd", "messageHeadline": "Handle initializers in subgraphs when inlining (#12758)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 25, "deletions": 0}}, {"node": {"oid": "4f244e48d59ab8a970b8cc17b52d73d5886508c7", "messageHeadline": "fix CalculateHash for Big Endian platforms. (#12752)", "author": {"name": "tvkai", "email": "97374290+tvkai@users.noreply.github.com", "user": {"login": "tvkai"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "b83ea3c2ff305069db779c4b5f980cfd38915207", "messageHeadline": "Address prefast static analysis warnings (#12756)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 14, "deletions": 16}}, {"node": {"oid": "27304d9082e1ef6c0f97c8c975bc108e729764f2", "messageHeadline": "gcc should not less than 7 (#12771)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3e57cd88fccdc0f0fa5902aec7054d01492829ae", "messageHeadline": "Csharp docfx update (#12755)", "author": {"name": "Cassie Breviu", "email": "46505951+cassiebreviu@users.noreply.github.com", "user": {"login": "cassiebreviu"}}, "additions": 16, "deletions": 15}}, {"node": {"oid": "80c8d934b84f0083558b61550f6180e6d8f42423", "messageHeadline": "Add debug option to packaging pipeline (#12685)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 26, "deletions": 8}}, {"node": {"oid": "817dc943457469ee962c2c2e67e22cfb8072ba26", "messageHeadline": "Add first pass of rocm kernel profiler (#10911)", "author": {"name": "mwootton", "email": "michael.wootton@amd.com", "user": {"login": "mwootton"}}, "additions": 1068, "deletions": 7}}, {"node": {"oid": "ee543a47f6bfa7061f5f25eb5390d8cab86be91f", "messageHeadline": "upgrade cuda version on ci pipelines (training CI pipelines) (#12708)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 20, "deletions": 15}}, {"node": {"oid": "64e88061486d062fde1a8a4ab3a87c9ab95f51de", "messageHeadline": "Address some static analysis warnings.", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 32, "deletions": 16}}, {"node": {"oid": "c270ea148a934136e5bcb8a3ad3b5c5be3bf4bde", "messageHeadline": "Move 'using common::Status;' from common.h to status.h.", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "3ff75fa05fff7582d43521cb6189bebe7adbb099", "messageHeadline": "Address static analysis warnings (#12711)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 11, "deletions": 7}}, {"node": {"oid": "34d90dd5bd1e14c31e78a3dad8927ee5b2a5f417", "messageHeadline": "mac-objc-static-analysis-ci-pipeline increase timeout (#12737)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "c9fd193ef6af1522bd82fc3d22e616588b1832c1", "messageHeadline": "Make TRT EP fully support control flow op and its subgraphs (#12692)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 212, "deletions": 7}}, {"node": {"oid": "a972db06bf33dcf49408b06c5911be97934856fc", "messageHeadline": "Disable SYMMQGEMM benchmark for CPU other than ARM (#12739)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "5bdb1d414632cf8930286ffdbd013b553c235460", "messageHeadline": "Add Tunable GEMM composed from rocblas and composable kernels (#12599)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 223, "deletions": 83}}, {"node": {"oid": "46c074a6c847d05e91863e5f335735d830dedb12", "messageHeadline": "Update composable kernel and enable experimental inter wave schedulin\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 19, "deletions": 14}}, {"node": {"oid": "3bb5fb0f90770a128f9055c2bde7f29a79ccbbd8", "messageHeadline": "moving training pipelines from cuda 11.5 to 11.6 and deprecating 11.3\u2026", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 7, "deletions": 222}}, {"node": {"oid": "f76b40aa5b71667fb12dda78f9c130145624ce4e", "messageHeadline": "Change TunableOp to use a type erased interface (#12597)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 496, "deletions": 453}}, {"node": {"oid": "baf141a0849c1941ad57a62fb022a2a2115b612c", "messageHeadline": "Enable xnnpack EP in Android AAR package (#12720)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "8483b9c6e3bb391525e5c8249bdc9ce0cc1a5522", "messageHeadline": "MacOS pipeline and MAUI CoreML fixes (#12724)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 7, "deletions": 11}}, {"node": {"oid": "ebff15d74394b8549a5af0f5f772073a928b5a3a", "messageHeadline": "Pinning manual seed (#12714)", "author": {"name": "abhi-ort", "email": "109558862+abhi-ort@users.noreply.github.com", "user": {"login": "abhi-ort"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "e85dce8ceac6dc2b2cf5dcf7ca019272d1ddfc2e", "messageHeadline": "Add csharp docfx (#12596)", "author": {"name": "Cassie Breviu", "email": "46505951+cassiebreviu@users.noreply.github.com", "user": {"login": "cassiebreviu"}}, "additions": 61853, "deletions": 12}}, {"node": {"oid": "5104c7dbd33a8049e6dd5016c5e8c22b8f2e8da9", "messageHeadline": "Fix Prefast Warnings (#12717)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 25, "deletions": 34}}, {"node": {"oid": "5be3e87c717c280a2d5662e6545035e70cbbbc6d", "messageHeadline": "[js] upgrade minimist@1.2.6 (#12689)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 25, "deletions": 25}}, {"node": {"oid": "cde504ebbf1a9b801313c2ec61448abca607c77d", "messageHeadline": "Fix/Suppress some VC static analyzer warnings (#12713)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 20, "deletions": 4}}, {"node": {"oid": "dee2fdffb00247aa15bcf5218738f81abb859b51", "messageHeadline": "Remove debug build/test in Mac CPU training (#12698)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "d91f017da1d19bb5f68e22372400488fe79164df", "messageHeadline": "remove redundant publish unit test results (#12697)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 7}}, {"node": {"oid": "eba4f77d005b993ca493c3b085dc8a7886405d47", "messageHeadline": "enable xnnpack in default_full_aar_build_settings (#12682)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "f1528ea50f618bbdc9ad7ce54eac52d7061505ed", "messageHeadline": "Fix arithmetic overflow warning. (#12712)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7927d525a7fd434d46e4a9be0ab7871b006a60b9", "messageHeadline": "Remove CUDNN path from CI build scripts (#12671)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 79, "deletions": 33}}, {"node": {"oid": "3f47119f3330a77b8d4e071dd3cf5f150119e095", "messageHeadline": "DML EP Fix InstanceNormalization with 3D tensors (#12693)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 30, "deletions": 10}}, {"node": {"oid": "94f76b944ec701ddacd47f1c8e47373a98177821", "messageHeadline": "nightly pipeline build using PTCA image. (#12605)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 31, "deletions": 0}}, {"node": {"oid": "0757d5133428a9c0afb65bc0f74b3ced81d77cd4", "messageHeadline": "Fix Java api docs broken link (#12686)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "53ecb9e6355dec171ed61d6daae13d89d53407bd", "messageHeadline": "Update Supporting DS Version to 0.7.1 for ORTModule (#12696)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "de3d772995cee40b57d932b91fa3818ba1986b42", "messageHeadline": "Check GCC version (#12680)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "8d657de4b2a094c27bba8f37a29c9b2583d54e08", "messageHeadline": "Update Newtonsoft.Json version to 13.0.1. (#12691)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "73e5741a9a904249de0ad5e7e659a9dc2ba34408", "messageHeadline": "Enabling softmax grad and logsoftmax grad on ORT (#12614)", "author": {"name": "abhi-ort", "email": "109558862+abhi-ort@users.noreply.github.com", "user": {"login": "abhi-ort"}}, "additions": 87, "deletions": 2}}, {"node": {"oid": "cb2601c5ea54f491bc8934d401fe041f498bf132", "messageHeadline": "Update mac-ci.yml to increase macOS build jobs' timeout value to 3 ho\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8d78f96dfed3d1b51973d89f722f41ba0389fb9b", "messageHeadline": "[CUDA] Fuse add bias and transpose into one kernel in Attention (#12670)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 30, "deletions": 14}}, {"node": {"oid": "6246662b1d1b0ba33a9eebfcc426397f62335f82", "messageHeadline": "[Dup] Fix SAME_UPPER/SAME_LOWER (auto_pad attribute) in ConvTranspose\u2026", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 13, "deletions": 22}}, {"node": {"oid": "c144acc5344e98e07bc6e07f013701a5e05e710d", "messageHeadline": "Replace 'master' branch ref to 'main' in the code (#12547)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 131, "deletions": 132}}, {"node": {"oid": "d93e6533b7bb9ba32bd5c4f8cc976ec37b838de1", "messageHeadline": "Format bert or transformers code (#12646)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1371, "deletions": 945}}, {"node": {"oid": "dc486d146b125db780fa2b71eda16875d01cd5e0", "messageHeadline": "Make ORT callable from various Pytorch compilers (LazyTensor, TorchDy\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 2263, "deletions": 44}}, {"node": {"oid": "53090f620ee7e84f79eabf5583a1a20049ab10f2", "messageHeadline": "Fix attribute renaming bug in function inliner (#12445)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 33, "deletions": 1}}, {"node": {"oid": "a078c8d99beb7fe5b620c4eb5264248f5247c368", "messageHeadline": "Update Supporting Deepspeed Version of ORTModule's FP16_Optimizer (#1\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8456f5fd97bf9e3b48cc4edc6013dd49b24ab98c", "messageHeadline": "qdq_util bug fix (#12647)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "2102b8f67cefa288af6f21eb34e68f63631a9b18", "messageHeadline": "Avoid duplicate symbol error between ONNX and ORT for ostream operato\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 42, "deletions": 91}}, {"node": {"oid": "f40e90c33f9098aaf90948fb469adc131349871f", "messageHeadline": "[js/web] fix incorrect shader for 'Resize' (#12588)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "bfdd191eecb2a51c8d626d5bfa1b32351c33ee37", "messageHeadline": "[wasm] use same export name for SIMD/NOSIMD build (#12545)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "aa85092b51ced97507d7fec12c4c67d6d376aac7", "messageHeadline": "DML EP squeeze all axes when empty (#12649)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "b270334e1e8365ba00f2db8aca1150af827bab16", "messageHeadline": "Update numpy version from 1.21.0 to 1.21.6 to avoid building it from \u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "56dd0176a172f18074deb1f8afb909fa05d39e3a", "messageHeadline": "QDQ debugger - Adding Error Calculator (#12632)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 76, "deletions": 3}}, {"node": {"oid": "81b128b5e97284f576ed8aaf18404980cb45fe34", "messageHeadline": "Qlinearsoftmax take FLOAT lookup-table (#12574)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 49, "deletions": 38}}, {"node": {"oid": "82b724fa5e8a15a94dc32090ab3cddd347b15af0", "messageHeadline": "[oneDNN] Improve DequantizeLinear operator performance. (#12611)", "author": {"name": "Erick Mu\u00f1oz", "email": "erick.munoz.alvarado@intel.com", "user": {"login": "eralmual"}}, "additions": 124, "deletions": 15}}, {"node": {"oid": "d1ba8015701bf4fd27a2d351497a63668ce7092c", "messageHeadline": "Add BuildError for --gen_doc and --enable_training (#12630)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "9481893b580f0e025cafccd9a0f158174847250a", "messageHeadline": "Replace to lock_guard as lighter class for locking (#12616)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "f2db6bb29341edf28e37c3456458a10023d2b261", "messageHeadline": "weight matching (#12607)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 258, "deletions": 63}}, {"node": {"oid": "8a038b9b0c313317318e6a362520a013d3db95a1", "messageHeadline": "Fix a build error (#12600)", "author": {"name": "Haoming Chen", "email": "haominghoward@gmail.com", "user": {"login": "hmchen-github"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ce01ed02da48cf25317eb2d4dfa4841012bf122c", "messageHeadline": "Improve LongformerAttention performance: AddBiasTranspose and New wei\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2039, "deletions": 879}}, {"node": {"oid": "7df2e8c5ccbe6d05ebdddd59b971358fd14c48b4", "messageHeadline": "Refactor with std::variant (on device training) (#12383)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 227, "deletions": 246}}, {"node": {"oid": "caabfcd9203ccf70e52aaeded315269efda61652", "messageHeadline": "Replace references to onnxruntime 'master' with 'main' in Dockerfiles\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 27, "deletions": 29}}, {"node": {"oid": "9d10badc552d0e77b93b08d9e598006bef2843c1", "messageHeadline": "Add build option to link TensorRT prebuilt parser (#12602)", "author": {"name": "yf711", "email": "109183385+yf711@users.noreply.github.com", "user": {"login": "yf711"}}, "additions": 48, "deletions": 19}}, {"node": {"oid": "733db31420568fd47468eb9c7bd879dddb63f92c", "messageHeadline": "[Java] JNI refactor for OrtSession (#12496)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 466, "deletions": 300}}, {"node": {"oid": "eb6aa861cfa7295ee9f7145db44aaec708e8ce1c", "messageHeadline": "QDQ debugger - activations compare (#12544)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 177, "deletions": 42}}, {"node": {"oid": "30ee5a4f79616c6a02b5a84136c9b6467b1734d5", "messageHeadline": "release calibrator before deleting temporary files (#12601)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 104, "deletions": 0}}, {"node": {"oid": "19a9690885ee962876b300c7b85e1b79129f01c8", "messageHeadline": "ppc64le: fix MlasQLinearMulKernel's VSX code to work with inputs of 3\u2026", "author": {"name": "Maxiwell S. Garcia", "email": "maxiwell@linux.ibm.com", "user": {"login": "maxiwell"}}, "additions": 25, "deletions": 73}}, {"node": {"oid": "616677104a0b16461c3958e3952017b423cc97fb", "messageHeadline": "ONNX Protobuf natvis with some google::protobuf (#12580)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 300, "deletions": 0}}, {"node": {"oid": "f5e3517c39b624bd027fdd9f5d3146d535bd74ac", "messageHeadline": "Add Learning Rate Scheduler C API (#11957)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 177, "deletions": 11}}, {"node": {"oid": "73da3f3705003bf4e4d872047a4fabb8c1856243", "messageHeadline": "Add TRT uint8 support (#12570)", "author": {"name": "Kevin Chen", "email": "45886021+kevinch-nv@users.noreply.github.com", "user": {"login": "kevinch-nv"}}, "additions": 20, "deletions": 0}}, {"node": {"oid": "95f2a3e7e0adfdd858333313ea29d8b3eb1e5271", "messageHeadline": "[js/web] update branch name for pull:wasm (#12548)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "cc9b3e1c374554b86e08a8b87d6787e7e36467b9", "messageHeadline": "Automate generation of javadocs and create PR with changes (#12515)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 48, "deletions": 0}}, {"node": {"oid": "0b0c51e02890ea35d0e5023681f2362421e44ceb", "messageHeadline": "Support direct usage of ORT format model flatbuffer for initializers \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 195, "deletions": 72}}, {"node": {"oid": "bc353c7afea850164ecd9e2dab4a782c91dfc1a8", "messageHeadline": "Add FusedConv Op to ROCm (#11792)", "author": {"name": "Xinya Zhang", "email": "Xinya.Zhang@amd.com", "user": {"login": "xinyazhang"}}, "additions": 460, "deletions": 10}}, {"node": {"oid": "eb827bd3e5cf4110023ccebaf1789ce5545561ec", "messageHeadline": "[ROCm] NGramRepeatBlock, LongformerAttention and DecoderAttention Ops\u2026", "author": {"name": "Xinya Zhang", "email": "Xinya.Zhang@amd.com", "user": {"login": "xinyazhang"}}, "additions": 620, "deletions": 94}}, {"node": {"oid": "95df5dac5174464b870b9a4e741c33bc012232f1", "messageHeadline": "do not quantize Relu/Clip if their inputs are not quantized (#12565)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 82, "deletions": 11}}, {"node": {"oid": "67f6b7ce293abb7ef57e9d5dda1d86e839d10ff4", "messageHeadline": "DirectML GEMM broken in opset 11 and 13 when optional tensor C not pr\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "580f2294bc50556a8e70cbaa02d39c595442efed", "messageHeadline": "Adding w_zero_point to conv_integer_test.cc (#12423)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 53, "deletions": 45}}, {"node": {"oid": "3d009cdde3f60144e877f69408d325e39e931014", "messageHeadline": "Updating binary ops in eager mode to support broadcasting. (#12560)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 348, "deletions": 41}}, {"node": {"oid": "24eab921bee4dbf9002942fbc514b2b2d9ee3a64", "messageHeadline": "Enable PythonOp for --enable_training_torch_interop build (#12539)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 107, "deletions": 45}}, {"node": {"oid": "b59ccbc75b6331a72caa93078e1b52627fe00574", "messageHeadline": "Add big endian support to murmurhash3 (#12549)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 38, "deletions": 13}}, {"node": {"oid": "018fba9b7463664e14e4038b10545a4eb6037516", "messageHeadline": "Fix Compile Warning (#12552)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 4, "deletions": 13}}, {"node": {"oid": "ac7538b909990f89f9bfebb325cd53c3d31cf158", "messageHeadline": "Remove CUDA 10.2 support (#12541)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 23, "deletions": 95}}, {"node": {"oid": "819c36701f066e2e37a86fa7c4a169ec80b0e374", "messageHeadline": "[xnnpack]  basic QDQ operators support (#11912)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1981, "deletions": 230}}, {"node": {"oid": "3e78f3cf1f458f56652c0e4444b34658bd11b3b6", "messageHeadline": "Add win-ci pipeline for on-device training (#12513)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 40, "deletions": 17}}, {"node": {"oid": "b2382dc43a92910c0e57fd4bfc31ed8649cd8521", "messageHeadline": "fix qdq relu removal bug (#12542)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 97, "deletions": 4}}, {"node": {"oid": "c10704a50113a2c8083afddc3d81ddde1ae27149", "messageHeadline": "Use alignas instead of naive padding to avoid false cache sharing (#1\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 23, "deletions": 6}}, {"node": {"oid": "25032f1756228e5265b1ef6bc3addb1d88c3e498", "messageHeadline": "Add default to TRT datatype switch statement (#12533)", "author": {"name": "Kevin Chen", "email": "45886021+kevinch-nv@users.noreply.github.com", "user": {"login": "kevinch-nv"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "c0d396d1769c968bad52fc3f4afb56488a1dbebc", "messageHeadline": "Restrict \"Component Detection\" task to Lotus project only (#12536)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 8}}, {"node": {"oid": "e810480403ebdd5f3438431ef2fc060c70a52b66", "messageHeadline": "Replace the occurrences of  \"master\" to \"main\" in yaml files (#12534)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 18, "deletions": 35}}, {"node": {"oid": "64e991a9fc044004859913a7e5d0940e93f703ee", "messageHeadline": "[Qlinearsoftmax] contrib cpu (#12177)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 1014, "deletions": 6}}, {"node": {"oid": "0c6037b5abe571fc43a55ef7a9d2f846820fbe5d", "messageHeadline": "Bugfix for BiasSoftmax Fusion (#12517)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 46, "deletions": 5}}, {"node": {"oid": "0d9a02e64775c58080641e00fb1139dc3c687dcd", "messageHeadline": "Eager Mode - Support Concatenation via aten::cat.out (#12527)", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 72, "deletions": 0}}, {"node": {"oid": "47b787c28fd1af4a86c1b189f3ef2b9675d9bb10", "messageHeadline": "Python module for dumping activation tensors when running an ONNX mod\u2026", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 329, "deletions": 7}}, {"node": {"oid": "2681648f5bde61b2f784b3115586631f3fcc9b28", "messageHeadline": "Load checkpoint in cpp (#12352)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 66, "deletions": 41}}, {"node": {"oid": "ee3b75749264e6276c01b7c688e9a8ddd1e73eec", "messageHeadline": "Add codeowners for requirement files (#12512)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "2bed0d4abb33178da7361213f68ba68939a39a95", "messageHeadline": "[CUDA] SoftmaxCrossEntropy Kernels Refactor (#12482)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 304, "deletions": 444}}, {"node": {"oid": "cfa09d16d9c588fda79a2dddd65f3927ad0a9d6f", "messageHeadline": "[CUDA] Mod Op Kernel (#12499)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 195, "deletions": 3}}, {"node": {"oid": "a2dc3e9eacdfd67e3e4538d95a26bd52685e5a2e", "messageHeadline": "Improve the compilation speed when compiling for multiple architectur\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 32, "deletions": 5}}, {"node": {"oid": "56bd96a3f5a23d1fe3d758ef8ad68c8fab3f224c", "messageHeadline": "Incrementally free initializers while saving to OrtValue instances (#\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 40, "deletions": 22}}, {"node": {"oid": "730240d2a5a888523a74a92102fe3c57513a5a39", "messageHeadline": "remove the link the comments (#12510)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "8a86b346a5fef042ab8a91a50eb05feab482e122", "messageHeadline": "[Java] JNI refactor for ONNX Tensor (#12281)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 277, "deletions": 204}}, {"node": {"oid": "8c5c2834711092bb864690c2a8bcfc0d1d9f710b", "messageHeadline": "new quantized operators split (#12495)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 155, "deletions": 2}}, {"node": {"oid": "9c055770214048b5b25c61feb9daef234809b372", "messageHeadline": "Fix various warning in kernel explorer (#12501)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 9, "deletions": 13}}, {"node": {"oid": "bdd6b00c9a6a5cdd2925762a9058e5c46f4a787f", "messageHeadline": "set zero point to 0 if all value are 0.0 (#12470)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 45, "deletions": 4}}, {"node": {"oid": "ddea1e48df47a3467141e360b828a41426336cd4", "messageHeadline": "Avoid false-positive dependent name lookup error by not depending on \u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "eb90b52a75c5dcb6132c82a65afa7b14ab750498", "messageHeadline": "DML EP fix training build error (#12461)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "e85e31ee803555f7b2eec4f7b11a1d8645b2c115", "messageHeadline": "Update ORTModule Default Opset Version to 15 (#12419)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 77, "deletions": 83}}, {"node": {"oid": "a7d62907743387003e0c341e8a22b232f6898a70", "messageHeadline": "CUDA kernel for ClipGradNorm for TensorSeq gradients (#12412)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 329, "deletions": 3}}, {"node": {"oid": "3e1b0ac4b3c6bf1c222f004631baa0eb3faf81b1", "messageHeadline": "[DELETE] delete python package rocm4.3.1 (#12480)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 7, "deletions": 229}}, {"node": {"oid": "b879dca51ca1431eddcc0b21eb360abf426e48fa", "messageHeadline": "Fix Python Packaging CI (Rocm) (#12477)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "8d830adf249024d60d58ced349831c507781b007", "messageHeadline": "Rework parts of Graph::Resolve to reduce memory usage (#12176)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 78, "deletions": 100}}, {"node": {"oid": "f39354d7cb16d5540dc099b0f50417ce0ddea21f", "messageHeadline": "Add composable kernel GEMM baseline for kernel explorer (#12364)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 550, "deletions": 167}}, {"node": {"oid": "37995a7245895c1116e67a90d72c4c568dac8282", "messageHeadline": "[CUDA] BiasSoftmax Supporting New Pattern (#12361)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 485, "deletions": 1089}}, {"node": {"oid": "d452462b5e45a26e84691c1e19260d4a25568fb9", "messageHeadline": "Lironkesem/unsqueeze_and_squeeze (#12421)", "author": {"name": "LironKesem", "email": "52330564+LironKesem@users.noreply.github.com", "user": {"login": "LironKesem"}}, "additions": 60, "deletions": 0}}, {"node": {"oid": "a4ef0e7f7b6628b473f002bfc6368b11b9a09132", "messageHeadline": "Remove dynamic allocation for ThreadPool ParallelSection (#12429)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 25, "deletions": 38}}, {"node": {"oid": "ac10f33d2d7d6eccb1f8040e8ca7fd1ca76417d4", "messageHeadline": "Enable quant op to share quantization parameter between input and oup\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 391, "deletions": 822}}, {"node": {"oid": "52d46997880348c7f695b3ec1f8e1e1084bfb699", "messageHeadline": "Minor doc fixes (#12388)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "3efd9a73bb783da7f52ab61c84c166c3a4bbdbed", "messageHeadline": "Refactor InferenceSession Load member functions. (#12430)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 83, "deletions": 132}}, {"node": {"oid": "97268e023c7f1c3e573666483126a0b03a9c42ad", "messageHeadline": "dev notes for layout transformer (#12396)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 58, "deletions": 0}}, {"node": {"oid": "a3de1bbf7d97224edd019a347d0d9561b1cad291", "messageHeadline": "Update script to find optimizers that potentially need supported opse\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 49, "deletions": 8}}, {"node": {"oid": "77cab7a3a50bd6d0ecedbb99196dc4bcbaebed60", "messageHeadline": "[ROCm] Add AveragePool, GlobalAveragePool, MaxPool, GlobalMaxPool Ops\u2026", "author": {"name": "Xinya Zhang", "email": "Xinya.Zhang@amd.com", "user": {"login": "xinyazhang"}}, "additions": 123, "deletions": 42}}, {"node": {"oid": "d1497bdf62b16a929762cdc71012b38aa7dbcaf1", "messageHeadline": "[oneDNN EP] Optimized DynamicQuantizeLinear operator (#12403)", "author": {"name": "Erick Mu\u00f1oz", "email": "erick.munoz.alvarado@intel.com", "user": {"login": "eralmual"}}, "additions": 82, "deletions": 96}}, {"node": {"oid": "7f58bd72361ab7d9b141d8bd5e69757a5402e0ef", "messageHeadline": "Perform graph transformations during offline tooling (#12422)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 61, "deletions": 1}}, {"node": {"oid": "dc984a03d506f02c9cce4ccc460ef8c614ea5019", "messageHeadline": "Container and memory allocation guidelines (#12387)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 77, "deletions": 13}}, {"node": {"oid": "97a340bf4857d60b35b0e058cebaee6ecf86ea88", "messageHeadline": "Fix integer overflow in LongformerAttention (#12435)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "44ec2cf08870ea24cd2a43448b9878d31432b4e4", "messageHeadline": "Update publish-python-apidocs.yml (#12433)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 13, "deletions": 6}}, {"node": {"oid": "b622e5fa9baac1c7f9084a9e51b036c72c93e8fa", "messageHeadline": "Support vocab_mask/prefix_vocab_mask/no_repeat_number in greedysearch\u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 187, "deletions": 205}}, {"node": {"oid": "01f3a197d7ef0f35144be01f585f6670421e35ee", "messageHeadline": "[ROCm] InstanceNormalization, BatchNormalization and LRN Ops (#11972)", "author": {"name": "Xinya Zhang", "email": "Xinya.Zhang@amd.com", "user": {"login": "xinyazhang"}}, "additions": 248, "deletions": 41}}, {"node": {"oid": "99d2a63e1ac222ac5955018055166b6a22f6387a", "messageHeadline": "Set Fix Seed For SoftmaxCrossEntoryLoss Related UTs (#12432)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "26dc09417b549fd19aa3b853cab842aa1f1fee30", "messageHeadline": "[oneDNN ep] matmulinteger postop fusion (#12354)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 1656, "deletions": 53}}, {"node": {"oid": "5d610bc8eb1529b87d473f5fff3031ce390265dd", "messageHeadline": "Disable CG task in PR pipelines (#12426)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 17, "deletions": 29}}, {"node": {"oid": "feed5da435f9ffa2819bad9ef1894524c72c3cb1", "messageHeadline": "[js] loosen test timeout (#12427)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "54d5e86981fa5425a3d1cb6cc13bfa1605523010", "messageHeadline": "Add cast before copy for dissimilar scalar type (#12391)", "author": {"name": "smrkatte", "email": "105341386+smrkatte@users.noreply.github.com", "user": {"login": "smrkatte"}}, "additions": 21, "deletions": 13}}, {"node": {"oid": "c9e0d0f8b6e64e3dbfa2cb36823ed593ee4c75b2", "messageHeadline": "[js/node] upgrade terser version (#12351)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 137, "deletions": 45}}, {"node": {"oid": "1a64b94f604e987c03c7d616ef61fffcc26eb88b", "messageHeadline": "Fix a small issue in nuget packaging pipeline (#12405)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "eebaf5f2701e7d0417c4a06fbb824222aa384555", "messageHeadline": "Adjust and fixx abseil-cpp debugging visualization (#12415)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 17, "deletions": 6}}, {"node": {"oid": "ca6b4221fe4a824faf1e5c9344d82607b262e8d7", "messageHeadline": "[js] Bug fix - permission issue with ensureSymlinkSync (#12369)", "author": {"name": "shalvamist", "email": "94086448+shalvamist@users.noreply.github.com", "user": {"login": "shalvamist"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b39257a5e6ed6768d4872691ce3be5d6072928e6", "messageHeadline": "Enable support of multi-level nested control flow ops model for TRT E\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 39, "deletions": 4}}, {"node": {"oid": "de3a91d85d9a72ef4bd7f902afd9d4c4b85aefe8", "messageHeadline": "Revert TRT EP cache refactoring  (#12376)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 232, "deletions": 300}}, {"node": {"oid": "5d1173fe68d97612f61dd2a240fe871cae926cba", "messageHeadline": "Run IOS pipeline concurrently (#12400)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 17, "deletions": 2}}, {"node": {"oid": "63d64636f682a27fc651acadf4dd42ed54e39060", "messageHeadline": "Add the comment linking to wiki (#12398)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "315e00653211b81129b4a8718f74d35afd4a2aaf", "messageHeadline": " adding a comment on nll_loss_forward.output that can not be implemen\u2026", "author": {"name": "LironKesem", "email": "52330564+LironKesem@users.noreply.github.com", "user": {"login": "LironKesem"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "62922f4c3cee5363ba55dbbf0c9515ef05a5d900", "messageHeadline": "Eager Mode generator: add comments, rename functions (#12385)", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 70, "deletions": 76}}, {"node": {"oid": "f77ab4fea6220d803b6b54dcab84e2105a8d0f28", "messageHeadline": "Manually add optimization flag for Android Release builds. (#12390)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "6bb807ef74252d1cca1d69c0ec7680972ff5ff6f", "messageHeadline": "add cuda compute 8.7 to Cmakelists.txt to support Nvidia Orin devices\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "3f66297499d426e17791fb93b8e3a23e0b1429b6", "messageHeadline": "code clean (#12392)", "author": {"name": "Cheng", "email": "wejoncy@163.com", "user": {"login": "wejoncy"}}, "additions": 862, "deletions": 868}}, {"node": {"oid": "1a4868e5c4c4a270ad91036e36f2a03410c4c278", "messageHeadline": "[TVM EP] Hot fix of build on Windows of TVM EP with ipp-crypto (#12381)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "8b4ad77ea236d7acaaf5b0a2283656d232aee65d", "messageHeadline": "pipeline can use last run's artifacts (#12379)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 422, "deletions": 392}}, {"node": {"oid": "6d1eb9509ec26ce09318ca14cc060c9a93ba498c", "messageHeadline": "Refine gradient accumulation (on device training) (#12363)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 179, "deletions": 47}}, {"node": {"oid": "7b4ce0c1e155f36544bd7b136c6bc695f0ad04d8", "messageHeadline": "Delete the build scripts that were copied from manylinux project (#12\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1101, "deletions": 11238}}, {"node": {"oid": "d5a1c01b382768bd7edf7681a7c57667c48282da", "messageHeadline": "Add C++ Session ctor taking model bytes and OrtPrepackedWeightsContai\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 246, "deletions": 226}}, {"node": {"oid": "df8dd41a8e0eae652699f0d2782c5cf9b2d4d252", "messageHeadline": "Automatically run workflows to generate API docs PRs (#11749)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 17, "deletions": 17}}, {"node": {"oid": "9559d25da93557ba438ded713e37cf57ee7af48b", "messageHeadline": "ORT Eager Mode Generator - make smaller functions (#12371)", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 369, "deletions": 239}}, {"node": {"oid": "f1a04078e97b47f2766b72de49e8a8eb6e4f6220", "messageHeadline": "Update CODEOWNERS to add a line for yaml files (#12378)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "c579497134909521237fbc593c38bde973f08359", "messageHeadline": "Fix TRT custom op issue (#12283)", "author": {"name": "Yateng Hong", "email": "yatengh@microsoft.com", "user": {"login": "toothache"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "651406974970ebd9ee6788d5c091a3b7dde8a270", "messageHeadline": "Make memory profiler work with multiple session runs. (#12317)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 283, "deletions": 212}}, {"node": {"oid": "e4bd41fb3b41ea9062b1ebecf9db21925d835f41", "messageHeadline": "[ROCm] Enable Einsum for inferencing perf (#12360)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 477, "deletions": 3}}, {"node": {"oid": "9c0fa65110a1a4823d83cb0ea9865d4874539e19", "messageHeadline": "Scope CreateFileMapping2 to valid API partitions (#12374)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0e85af6990ac41b8925722d4aee990909421a729", "messageHeadline": "Add MAUI csharp\\sample\\InferenceSample\\ project (#12356)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 992, "deletions": 0}}, {"node": {"oid": "805aa297fccf0b872565cf6725fb6679772c5718", "messageHeadline": "Remove preview keyword from DirectML pacakge (#12368)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7a7e372b9f689704a5ca5f5efb4d381a026c6f1b", "messageHeadline": "Remove training cuda 10.2 pipeline (#12347)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 0, "deletions": 223}}, {"node": {"oid": "6e892a95b4ee1234add6b33d0172858da900a059", "messageHeadline": "Use specific Android NDK version in CI builds. (#12350)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 51, "deletions": 10}}, {"node": {"oid": "73919a67563fa6b2bf33524be4770f9abb5db97d", "messageHeadline": "Convert DQ node with const weight tensor int8 to uint8 (#12331)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 279, "deletions": 187}}, {"node": {"oid": "e2423bb55c4e3c2912e3e52ec4e1af883571d066", "messageHeadline": "[TVM EP] Build on Windows with ipp-crypto support (#12336)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 75, "deletions": 35}}, {"node": {"oid": "7a298c916a4b732b025d9310f59399ab76385b7c", "messageHeadline": "Add Pow-15 for LayerNorm Fusion and FastGelu Fusion (#12314)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "186ba6e9f2f59d59e860faf912a62763f3abc9a3", "messageHeadline": "[js/rn] upgrade package react-native@^0.69.1 (#12155)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1004, "deletions": 867}}, {"node": {"oid": "e6bb447101aaab9e14fc8a37a149183d4d01d63f", "messageHeadline": "Change native folder name for java macos arm64 (#12335)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "148b1efe5e698138f473bd500778e78d49ee31ca", "messageHeadline": "[js/web] add ConvTranspose2D to WebGL backend (#11990)", "author": {"name": "101arrowz", "email": "arjunbarrett@gmail.com", "user": {"login": "101arrowz"}}, "additions": 279, "deletions": 10}}, {"node": {"oid": "d2b25a7c1c24660d40480439a6b56ecad2447c1c", "messageHeadline": "Reduce CI noise from Python lint (#12270)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 12, "deletions": 3}}, {"node": {"oid": "9cf6912bba05dea82a2facbbed243b6363873dae", "messageHeadline": "Fix ORT Eager Mode to work with Pytorch 1.12 (#12323)", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 30, "deletions": 27}}, {"node": {"oid": "e2eeffeafb90fdb69d58c8a633ef9ced99c8d2d3", "messageHeadline": "Cosmetic fix to AttentionFusion (#12329)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1163294699365ad07f6e914f2be3770f29be0a23", "messageHeadline": "Fixing up some python warnings. (#12319)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 81, "deletions": 69}}, {"node": {"oid": "1dd65b9ae3a6627c10eb1a4c17abaefc95181393", "messageHeadline": "Use memory mapped data for external data initializers with Cuda fix (\u2026", "author": {"name": "kiennguyen94", "email": "hakien94@gmail.com", "user": {"login": "kiennguyen94"}}, "additions": 118, "deletions": 38}}, {"node": {"oid": "4df4471d5eab15e366a4634e34840ac8ef5bfd98", "messageHeadline": "add missing build_java in Android testing stage. (#12187)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 43, "deletions": 25}}, {"node": {"oid": "2b2367efbf3e76837ae764f0c260d6d34bec5628", "messageHeadline": "Fix orttraining-linux-gpu-ci-pipeline (fairscale dependency) (#12320)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "51a799802a9bc4f21ffc5ebd2979ccfefb376f2e", "messageHeadline": "Move initializers from subgraph to the main graph to reduce memory (#\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 50, "deletions": 4}}, {"node": {"oid": "f3dcbf539aa410f31a0e4bce1807f79331fe1c1f", "messageHeadline": "Checkpoint load inference (#12168)", "author": {"name": "Adam Louly", "email": "adamlouly3@gmail.com", "user": {"login": "AdamLouly"}}, "additions": 129, "deletions": 3}}, {"node": {"oid": "de57daaab055766f6a5fa0c4ef34318fe611174b", "messageHeadline": "Eager mode: binary ops more complete behavior and testing. (#12293)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 395, "deletions": 337}}, {"node": {"oid": "3e014a5e5d1a94f354a8a169ca02b7ab2668498c", "messageHeadline": "Fix C header to stop people accidentally copying the OrtApi by value \u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 22, "deletions": 23}}, {"node": {"oid": "c40f73ae0cb986fd5e57c54764643adffbe2e37c", "messageHeadline": "Remove aten::binary_cross_entropy_with_logits from ATen Fallback (#12\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 0, "deletions": 78}}, {"node": {"oid": "3bf614fd470d6fa37104c0b24014e83540cb8e6e", "messageHeadline": "Eliminate memory allocations per recent profiling (#12225)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 289, "deletions": 227}}, {"node": {"oid": "972bb9676c521b4c5957b8812f6872e56709fed8", "messageHeadline": "Bump terser from 5.7.0 to 5.14.2 in /js/common (#12248)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 137, "deletions": 45}}, {"node": {"oid": "ddb45e9126aa668102b0fa59c39e048577fdf2f1", "messageHeadline": "On device training CI pipeline (#11987)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 126, "deletions": 13}}, {"node": {"oid": "8d0e86dec85281db3623809a72e5f6a36ac04eb0", "messageHeadline": "Apply project formatting rules to ort_aten.cpp (#12294)", "author": {"name": "Jameson Miller", "email": "jamill@microsoft.com", "user": {"login": "jamill"}}, "additions": 280, "deletions": 292}}, {"node": {"oid": "0fa3aeb65c49311d732672a13ddf3e1c8b165fbe", "messageHeadline": "[CUDA] Add Strided Tensor Support for Expand->GatherElements for Trai\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 983, "deletions": 413}}, {"node": {"oid": "75bda9f267382e59393250344248438e82a4ad15", "messageHeadline": "CPU AdamW implementation (#11978)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 3626, "deletions": 330}}, {"node": {"oid": "564dc32304d73ba0d73aaa1c0fee6b76dd0fe90a", "messageHeadline": "Initialize generated tensor data in onnxruntime_perf_test. (#12275)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 45, "deletions": 2}}, {"node": {"oid": "89ac61f4d446e4a3a628203c31f5291a3a739c4c", "messageHeadline": "support gpt2 model with greedy search (#12068)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2024, "deletions": 398}}, {"node": {"oid": "cb351388d02cdf93cf2ad9808d2c2ef0ebc8d373", "messageHeadline": "Minor fixes (#12276)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 8, "deletions": 7}}, {"node": {"oid": "0264a9c29b88691d1e6b64432560705d36ac99a1", "messageHeadline": "Bump ort version number (#11948)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 26, "deletions": 47}}, {"node": {"oid": "a71f7d33395afb0c50bae744e959e0d181ec06a9", "messageHeadline": "[NNAPI EP] Add some support for MatMul with batch inputs (#12261)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 516, "deletions": 61}}, {"node": {"oid": "4f57da78cf3170d82c7195c9454407792e107918", "messageHeadline": "OrtModule fix pytorch version comparison (#12280)", "author": {"name": "Juan Paez", "email": "juanfranciscopaez@gmail.com", "user": {"login": "juanpaez22"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "feabafe58bdac491ee67519c286927ad7771ea79", "messageHeadline": "Fix memory consumption discrepancy (#12266)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 181, "deletions": 145}}, {"node": {"oid": "f2533d3084b6af933f60575a32d31e0ecf3f7174", "messageHeadline": "Remove two lines in the Dockerfile for Github Codespace (#12278)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "ceb76429dbca2e7f1a9aaa708d71a07992a7b2df", "messageHeadline": "Merge pull request #12056 from microsoft/bmeswani/merge-training_dev/\u2026", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 10149, "deletions": 63}}, {"node": {"oid": "7194ec1894c53dac7f25896269f24c14c2c664a5", "messageHeadline": "fix bug: output of Concat is quantized twice in qdq format (#12254)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 14, "deletions": 16}}, {"node": {"oid": "a18b0805138460db5bcac1a29fb0e6cf726cc1d2", "messageHeadline": "clean up calibration model (#12255)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 13, "deletions": 9}}, {"node": {"oid": "45c0be8a25f2e4f5629c5ceee15cc1c4ca5b6bf1", "messageHeadline": "Modify generator for eager to use all inputs for determining promote \u2026", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "30ac6e87fa297ab0571bcb088496fc5da0e080fd", "messageHeadline": "Bump terser from 5.10.0 to 5.14.2 in /js/web (#12253)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 117, "deletions": 31}}, {"node": {"oid": "eb3b49b6a193771533f040381b6f9186b0eb749a", "messageHeadline": "[CoreML EP] Remove batch=1 restriction in depthtospace op support (#1\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 3, "deletions": 9}}, {"node": {"oid": "108b860dc1c5003271a171c4fb384d2bf2d52289", "messageHeadline": "Add dev container / codespace configuration (#12256)", "author": {"name": "Jameson Miller", "email": "jamill@microsoft.com", "user": {"login": "jamill"}}, "additions": 88, "deletions": 0}}, {"node": {"oid": "03dfcb0e87b6c28a75f074e3e6105c714c40819f", "messageHeadline": "[ROCm] Enable int8 for MatMulInteger Op (#11776)", "author": {"name": "Xinya Zhang", "email": "xinyazhang@utexas.edu", "user": {"login": "xinyazhang"}}, "additions": 69, "deletions": 6}}, {"node": {"oid": "cbf08c7a7b275e5cce11838e0b12e88c3245b1c4", "messageHeadline": "Make GetTrainingApi as a part of the OrtApis, add Training API docume\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 536, "deletions": 298}}, {"node": {"oid": "3d2bcb33866694cedf558d34e8006268b678217f", "messageHeadline": "Use unregister_custom_op_symbolic to unregister torch symbolics (#12146)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 8, "deletions": 6}}, {"node": {"oid": "496618594f40d2114cf03720526970cfc4c4a9e5", "messageHeadline": "Update supported ops md for NNAPI/CoreML EP (#12245)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 11, "deletions": 4}}, {"node": {"oid": "7dc45bc31118ae45eeaa29b3b2785806660a2594", "messageHeadline": "Implementing aten::gt.Scalar_out and aten::lt.Scalar_out (#12181)", "author": {"name": "LironKesem", "email": "52330564+LironKesem@users.noreply.github.com", "user": {"login": "LironKesem"}}, "additions": 93, "deletions": 92}}, {"node": {"oid": "007ef427493052ca4a3e90b4560baa8c2985b98d", "messageHeadline": "Fix: Test coverage is undercounting and profiling errors (#12260)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "5066ef118509b332d5a500e0394e2b2ba83ffe58", "messageHeadline": "Fix a bug in beam search custom attention mask allocation (#12240)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 41, "deletions": 0}}, {"node": {"oid": "0c78b713527ce5b3c0b4d3cc08094d7ac8dfa62f", "messageHeadline": "prepare test folder from GitHub (#12220)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1019, "deletions": 306}}, {"node": {"oid": "568d08994f1abdba9623c406afc09c4f04374541", "messageHeadline": "fix test_optimizer.py (#12219)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 110, "deletions": 125}}, {"node": {"oid": "c72bb8aaa9759954e35d394ee39b0f578ca56c86", "messageHeadline": "[js/web] add OffscreenCanvas support to WebGL backend (#12159)", "author": {"name": "101arrowz", "email": "arjunbarrett@gmail.com", "user": {"login": "101arrowz"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "471dbfc250aca638b1bb1e86fcb00fb9f84d5590", "messageHeadline": "[NNAPI] Add int32_t as supported input data type and other minor gath\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 89, "deletions": 76}}, {"node": {"oid": "5651d91c3250828c52cee52025e41bc5ab7dbb77", "messageHeadline": "Fix onnx version comparison (#12223)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "43e1e89453119ecdfea4463ae742f669ea956f9d", "messageHeadline": "Update aarch64 building pool to aiinfra-linux-ARM64-CPU-2019 (#12243)", "author": {"name": "Jian Chen", "email": "cjian@microsoft.com", "user": {"login": "jchen351"}}, "additions": 26, "deletions": 6}}, {"node": {"oid": "424120d0fa0d6e8ff61c1764a6b353bea19330ef", "messageHeadline": "cpplint & Eager mode: refactor and add comments to empty_* functions,\u2026", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 84, "deletions": 73}}, {"node": {"oid": "72c689a502a5508760ff737250849925c23d7899", "messageHeadline": "[CUDA] Use dim3.z to Handle Large Input For GatherGrad (#12250)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 40, "deletions": 19}}, {"node": {"oid": "ebfd81e67e200fc22612e81fd0b9793a8e1fac5e", "messageHeadline": "Fix BiasGeluGrad bug (#12200)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 67, "deletions": 14}}, {"node": {"oid": "3cdc6d7775d2df2b836c5d39c2582daa2151b324", "messageHeadline": "[ORTModule] Bugfix of torch.chunk's Custom Symbolic when chunks==1 (#\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 35, "deletions": 28}}, {"node": {"oid": "a0074ba9bc1622594e0198b1f0a5667e23153799", "messageHeadline": "Add baseline gemm for kernel explorer (#12050)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 277, "deletions": 0}}, {"node": {"oid": "add631410a4b671d622a8bf5872e7da41aee7905", "messageHeadline": "[ROCm] Re-enable ReduceL1, L2 and related tests (#12209)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 10, "deletions": 302}}, {"node": {"oid": "9b6ef17c5f4f5eff58eebbd351bec2d09d405911", "messageHeadline": "Eager opgen support for in-place operations with variadic args (#12125)", "author": {"name": "Juan Paez", "email": "juanfranciscopaez@gmail.com", "user": {"login": "juanpaez22"}}, "additions": 75, "deletions": 45}}, {"node": {"oid": "5e2109f7ef5f64eb49ad4e0fdc80f369d9845572", "messageHeadline": "[ROCm] Enable GridSample Op. (#11969)", "author": {"name": "Xinya Zhang", "email": "xinyazhang@utexas.edu", "user": {"login": "xinyazhang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4f106d2b3b2f55f3bd74f0f6bcbbcd4c304b9231", "messageHeadline": "Eliminate unnecessary status lock acquisition in TP (#12196)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 19, "deletions": 12}}, {"node": {"oid": "972e5e73008a8af20eab47905d5acbbf00d1373b", "messageHeadline": "Improve symbolic shape inference in transformers tools (#12217)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 95, "deletions": 70}}, {"node": {"oid": "975bb56e8c559c4db4b6d0059b7b8862150bfb79", "messageHeadline": "Eager mode - argmax_out: set output tensor (#12233)", "author": {"name": "Jameson Miller", "email": "jamill@microsoft.com", "user": {"login": "jamill"}}, "additions": 95, "deletions": 10}}, {"node": {"oid": "555e88982f7f84aa1ac6c1eb3c786d7ca5cbfe95", "messageHeadline": "Fix GH issue 12208 (#12224)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "2cb642927b06df336f69da5a92f640a7c4fecea5", "messageHeadline": "Simplify get_docker_image.py  (#12166)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 102}}, {"node": {"oid": "0c319d6e94c1b8b651eb5318266576fa4570ff9c", "messageHeadline": "Exclude implicit inputs from dump of encoder feeds in beam search (#1\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "66978c7ef59eeca79634c644d78d118208218a39", "messageHeadline": "[TVM EP][CI] Added TVMso EP testing into CI (#12188)", "author": {"name": "Alexey Gladyshev", "email": "wotpricol@mail.ru", "user": {"login": "KJlaccHoeUM9l"}}, "additions": 245, "deletions": 54}}, {"node": {"oid": "4235ebc1615dd4877c6a531c505750374d4c40ff", "messageHeadline": "Add eager mode support for mm.out (matrix multiplication). (#12214)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 96, "deletions": 1}}, {"node": {"oid": "bb5bd08545b6b12f61851692f762ec46bbaa5980", "messageHeadline": "[ROCM] Navi21 fixes pr (#11368)", "author": {"name": "Michael Melesse", "email": "micmelesse@gmail.com", "user": {"login": "micmelesse"}}, "additions": 31, "deletions": 23}}, {"node": {"oid": "173bcdbc7189dd43f4e16029fb5ebcffee0c92bd", "messageHeadline": "[CUDA] Split/Concat Kernel Optimization (#12175)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 403, "deletions": 410}}, {"node": {"oid": "ced7c2deac958391414d2bbf951f86e2fc904b05", "messageHeadline": "[js/web] use windowed Chrome for perf mode (#12157)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 20, "deletions": 4}}, {"node": {"oid": "b81b6526083144aa0232c5172b7014606cbe20f0", "messageHeadline": "Add --disable_shape_inference option to optimizer.py (#12215)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 31, "deletions": 9}}, {"node": {"oid": "93229949d45b91ba8e495705d4e6db71f513e4db", "messageHeadline": "Fix bug where onnxruntime_USE_NCCL flag would default to ON (#12195)", "author": {"name": "Sean Murray", "email": "59740888+seanmurr1@users.noreply.github.com", "user": {"login": "seanmurr1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "17b84c78f7fe6d9de9ad3da8151586e7946f6ba0", "messageHeadline": "remove identity in transformers model graph fusion (#12194)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 19, "deletions": 4}}, {"node": {"oid": "4d38b84e268611023e1322d8e255039fb8ae16a6", "messageHeadline": "Add file mapping for windows platform. (#12183)", "author": {"name": "caoting-dotcom", "email": "71617901+caoting-dotcom@users.noreply.github.com", "user": {"login": "caoting-dotcom"}}, "additions": 163, "deletions": 0}}, {"node": {"oid": "09af4a7fdd23ab6da687ce6aad74d5a86bc2603a", "messageHeadline": "remove wrong placed libs (#12201)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "d31db1aa575292ec1d12b8f0cc0929070fd5f4cd", "messageHeadline": "[TVM EP][CI] Integrate TVM EP into ORT public CI on Windows (#12161)", "author": {"name": "Alexey Gladyshev", "email": "wotpricol@mail.ru", "user": {"login": "KJlaccHoeUM9l"}}, "additions": 37, "deletions": 0}}, {"node": {"oid": "52095fb042ebcb9f1e520d58b23fb9cec260aab3", "messageHeadline": "Fix line spacing/break issue, extend existing tests (#12191)", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 18, "deletions": 4}}, {"node": {"oid": "a2dc6d32fca1d13ffbdbc03672f5192375956763", "messageHeadline": "OnnxRuntime Eager: Implement log_softmax with ONNX Ops (#12190)", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 110, "deletions": 5}}, {"node": {"oid": "9bca8405aa8cd1be3aab5fb1bc7a374a37d80835", "messageHeadline": "bitwise_and ONNX support (#12189)", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 22, "deletions": 1}}, {"node": {"oid": "89bf6c9b5d12ee522abc52c1fcff1d2d1df83b2e", "messageHeadline": "Simple eager training models (#12180)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 124, "deletions": 20}}, {"node": {"oid": "fafb24142fb7d3aa0c6128c56ff49710ffb3cf39", "messageHeadline": "add comment to explain local scalar dense (#12179)", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "05c31a036d5a2932ac09fe367cfcfbb125e32fb2", "messageHeadline": "fixing positions for beam search gpt2 (#12156)", "author": {"name": "Viswanath Boga", "email": "44417868+viboga@users.noreply.github.com", "user": {"login": "viboga"}}, "additions": 79, "deletions": 83}}, {"node": {"oid": "9ebef91a6faf754d6820f23028c596e6687f5684", "messageHeadline": "Update eager Readme.md (#12170)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 44, "deletions": 11}}, {"node": {"oid": "7b53b223b8bc9bd86164ad919192f95cf01b952c", "messageHeadline": "[UPDATE] update AMD CI pipeline to Rocm5.2 with torch1.11 (#12162)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 48, "deletions": 49}}, {"node": {"oid": "a7eb9fe3acb20b0e974cd1bcef758e2615d843a4", "messageHeadline": "Remove Apex Dependency For Deepspeed FP16_Optimizer (#12077)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 242, "deletions": 9}}, {"node": {"oid": "5da1e5d36d00fbae851cb675f93fb0db44bc6ef7", "messageHeadline": "Eager mode: Fix some python warnings. (#12167)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 8, "deletions": 17}}, {"node": {"oid": "51f8456c4da7bedcbf1aa8e4445cbcbd541ea909", "messageHeadline": "ppc64le: Optimizing the MlasQLinearMulKernel() to use VSX instruction\u2026", "author": {"name": "Maxiwell S. Garcia", "email": "maxiwell@linux.ibm.com", "user": {"login": "maxiwell"}}, "additions": 141, "deletions": 0}}, {"node": {"oid": "040c2f4517775ce8362903837346213babbdf6b1", "messageHeadline": "x86/64 U8S8 Gemm Precision Fix (#12088)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1325, "deletions": 6}}, {"node": {"oid": "48647bc7d788699b5eef35bcaea35f320263cb22", "messageHeadline": "Fix NonZero eager impl. (#12143)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 89, "deletions": 19}}, {"node": {"oid": "3b0aaa9e0e8a26943dfe3e07139c24bfc4fa68d5", "messageHeadline": "[TVM EP] support build on Windows (#11851)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 240, "deletions": 51}}, {"node": {"oid": "75cf5dc2c9260e082172ea4e8c5384aef6e4cfe2", "messageHeadline": "Fix GH issue 12151 by using inverse perms for updating DQ axis attrib\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 41, "deletions": 2}}, {"node": {"oid": "785f74979b52da8360b8ad416839553dadc71c44", "messageHeadline": "Rework cmake for kernel_explorer (#12079)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 242, "deletions": 78}}, {"node": {"oid": "5579d81fc885e831e2bac5dc83db47517685ad95", "messageHeadline": "[add] Add operator gemmfastgelu for ROCM (#12101)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 645, "deletions": 133}}, {"node": {"oid": "a9d0d3323e60bd8e4d89eb608e7195b27504d2be", "messageHeadline": "Use updated symbolic_helper.check_training_mode (#11900)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 14, "deletions": 3}}, {"node": {"oid": "178a413ca1ae74c3b7c70ee6c07fb86ead61b3c1", "messageHeadline": "List 3.10 as supported python version and remove 3.6 (#12141)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e0ed9f0f2fc6bd30b1e2302f2b0ed621e0e37f74", "messageHeadline": "[java] First part of the JNI error handling rewrite (#12013)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 432, "deletions": 306}}, {"node": {"oid": "a6fd1a3b85d4f0524b5e55f21a832b08dd0d73f6", "messageHeadline": "Eager mode generator improvements for multiple onnx operators and ext\u2026", "author": {"name": "msftlincoln", "email": "107071614+msftlincoln@users.noreply.github.com", "user": {"login": "msftlincoln"}}, "additions": 36, "deletions": 17}}, {"node": {"oid": "6e051016c1894775d78cd3873fc79c8cbd5e4edb", "messageHeadline": "Add Python package to perf test pipeline. (#12135)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 64, "deletions": 45}}, {"node": {"oid": "9647a3be40c78297cd6897a659249a72b85758f1", "messageHeadline": "Add tests for all unary aten ops supported in eager mode (#12087)", "author": {"name": "LironKesem", "email": "52330564+LironKesem@users.noreply.github.com", "user": {"login": "LironKesem"}}, "additions": 128, "deletions": 31}}, {"node": {"oid": "73310b2a0f6d623e95ae37d2ab4d7892a4654dc9", "messageHeadline": "Fix Reduced Ops build pipeline (#12144)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "c675c4750aee8f4172984422af02944d83588c3e", "messageHeadline": "include coreml_provider_factory.h in macos build instead of coreml_ex\u2026", "author": {"name": "Carson Swope", "email": "carsonswope@users.noreply.github.com", "user": {"login": "carsonswope"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "742f843efcc73739a915858051e187e04f2dcdb7", "messageHeadline": "RoiAlign CPU EP add warning for max mode with samples != 1 (#12136)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "f1047e04564cc59cb41b5590a3b830bcfbb753ef", "messageHeadline": "Fix minor python and cpp  warnings from previous PR. (#12140)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 13, "deletions": 7}}, {"node": {"oid": "99a370dd02dcf40e40062f92f09918edb421c4ec", "messageHeadline": "Update readme for OVEP  (#12122)", "author": {"name": "Preetha Veeramalai", "email": "preetha.veeramalai@intel.com", "user": {"login": "preetha-intel"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "418cfdc766337e6ef264e5ea9d84f8851245e899", "messageHeadline": "Update create_ort_attribute to set the tensor dimension and value cor\u2026", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 73, "deletions": 7}}, {"node": {"oid": "1c39d22f4ec32f120b5c7ed4b79ac2b99d14db91", "messageHeadline": "[ADD]  Rocm5.2 for Rocm python packaging pipeline (#12129)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 234, "deletions": 20}}, {"node": {"oid": "c6732c079b4fbf644bc939729c87d70195251d83", "messageHeadline": "pin protobuf version to be compatible with onnx (#12132)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d45c1a144e77d5901a3126f3dba23aa626d53cba", "messageHeadline": "[js/rn] Support UINT8 type for onnxruntime-react-native on Android (#\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 55, "deletions": 2}}, {"node": {"oid": "c04afae9a9f274330102f83240eb5295bfa01f20", "messageHeadline": "Add eager ops for unary ops with out. (#12106)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 56, "deletions": 32}}, {"node": {"oid": "2dd69cc3d9f90ce754a5711155af946b9e110d38", "messageHeadline": "Prevent unbounded growth of command allocator memory (#12114)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 12, "deletions": 18}}, {"node": {"oid": "3ce25db7ebb42ff7f803e842e6ca9ff6c3534953", "messageHeadline": "[js/rn] optimize exception message on Android (#12113)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 57, "deletions": 53}}, {"node": {"oid": "b50239251de32906e3123d0c0bc430c7a528d611", "messageHeadline": "[FIX] Add required variable for Rocm packaging ci pileine  (#12118)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "a9b9c7f69f8804f4390ee22008d3c21a40a22f53", "messageHeadline": "Add autotuning support to FastGelu (#12093)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 448, "deletions": 256}}, {"node": {"oid": "dbcf54aa41876b9ffc2aab3368dbce2f7e962c34", "messageHeadline": "Add hipified SkipLayerNorm code for ROCmEP (#12107)", "author": {"name": "Hubert Lu", "email": "55214931+hubertlu-tw@users.noreply.github.com", "user": {"login": "hubertlu-tw"}}, "additions": 535, "deletions": 1}}, {"node": {"oid": "97b03fedfff0e272a58bd61b378698a0d0823228", "messageHeadline": "check consumers of dq node before swap dq and transpose (#12099)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 62, "deletions": 13}}, {"node": {"oid": "446f899fed531ae7c3332a3678b7b39cb75cd250", "messageHeadline": "[ROCm] Temp disable AMD UT (#12105)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 18, "deletions": 17}}, {"node": {"oid": "bd76e21fb3ad2fc2190c702fe1c80551470af3b5", "messageHeadline": "Add pipeline for building perf test binaries. (#12067)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 16, "deletions": 0}}, {"node": {"oid": "1948b7c726c3285a0606cbf714ee2db183549b47", "messageHeadline": "Add eager support for eq and ne ops. (#12031)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 107, "deletions": 43}}, {"node": {"oid": "07b0469a23248410bf92fb0bfc8f705e50f51ac4", "messageHeadline": "Fix unused function warning for decodeMIDR(). (#12069)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 397, "deletions": 364}}, {"node": {"oid": "835ecb264d3027b70a8d0cab9cb04ba190e15f46", "messageHeadline": "Leverage vectorized load/write for SkipLayerNorm  (#11803)", "author": {"name": "Hubert Lu", "email": "55214931+hubertlu-tw@users.noreply.github.com", "user": {"login": "hubertlu-tw"}}, "additions": 277, "deletions": 101}}, {"node": {"oid": "7b8f45dd609824010d072dcc5906570f6f626807", "messageHeadline": "[ROCm] Enable build option for autograd (#11945)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 62, "deletions": 134}}, {"node": {"oid": "32a8751dc47ab9e0155cdc91243d8cc246c3e131", "messageHeadline": "DML EP Update to DML 1.9 (#12090)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 5, "deletions": 7}}, {"node": {"oid": "3446a3750cde99eda6c1fcac6e500451d4619855", "messageHeadline": "generate quantization parameter for outputs (#12089)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 49, "deletions": 41}}, {"node": {"oid": "3d53ce5e7b9db5b9595f420ab1d3087742e27ff4", "messageHeadline": "Remove ROCM macro for InPlaceAccumulatorV2 test", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "1aa27e127ce3bbbfe0ed4cde3ddac6c35688038c", "messageHeadline": "Resolve build conflicts with master", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "479e71a7a856c16884c70f49ce9e123098ff8094", "messageHeadline": "enable the extensions custom build for java and android (#11823)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "c20cbf0c97c9596c789fe861eb979d27511b4e2a", "messageHeadline": "Add undocumented attribute to disable generation of Java bindings fro\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ddb6202df7323d763de7781f0de3c74c64729e1b", "messageHeadline": "Add op tuning functionality and an example for vector add. (#12060)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 251, "deletions": 38}}, {"node": {"oid": "df712d80ca33b56ce7d63ca3acdb6a4888638c58", "messageHeadline": "Add data type check in ConvAddRelu fusion (#12058)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 97, "deletions": 4}}, {"node": {"oid": "57ac3d0a612e6117b094fb7ead63658c6f61a3d4", "messageHeadline": "Disable DML command list reuse for Xbox (#12063)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "4a335e979f07ec85e26cfe0dd6062b13edb99fe1", "messageHeadline": "Add kernel def hash for inplaceaccumulatorv2", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "ae88f435505ec3db7b3c378fefa1f006657637e5", "messageHeadline": "Eager mode: structure for supporting out= operators (#12066)", "author": {"name": "Jameson Miller", "email": "jamill@microsoft.com", "user": {"login": "jamill"}}, "additions": 114, "deletions": 3}}, {"node": {"oid": "8dc8d44087abc495e8c1a37a112d39398fff908e", "messageHeadline": "remove --disable_iobinding for trt ep benchmark (#12053)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "043816f895c5a0b6aeae39b062d16d5c102d6ebf", "messageHeadline": "Make C# runtest.sh automatically set latest opset (#12039)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "02b9b121272a5cd4f57a245eb15a2ca67c9d97ea", "messageHeadline": "Fix DML custom operators which set descriptor heap to command list (#\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 18, "deletions": 8}}, {"node": {"oid": "bfe1eca10c607bfcbc005889d91b5db1d8346686", "messageHeadline": "Add targets files for new .net6 frameworks (#12016)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 155, "deletions": 42}}, {"node": {"oid": "3e6b8d159aa8d93e2ad64ba55c7bd0311fb941b9", "messageHeadline": "Eager mode: implement resize_ operation (#12004)", "author": {"name": "Jameson Miller", "email": "jamill@microsoft.com", "user": {"login": "jamill"}}, "additions": 183, "deletions": 0}}, {"node": {"oid": "b858c2f7251aaef76b6bbd2d58e9d30979f7e422", "messageHeadline": "Extend lifetime of KernelDef when creating a standalone op (#12057)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "a457ddc41d8335997604f9d3bd44d14267b29a55", "messageHeadline": "Merge branch 'master' of https://github.com/microsoft/onnxruntime int\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 27129, "deletions": 13037}}, {"node": {"oid": "2e27a7e3305ab90d16786388ef505ad18a55fa6c", "messageHeadline": "Skip Constant Folding for ops producing an optional type output (#11839)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 20, "deletions": 5}}, {"node": {"oid": "0fa2041f6892af5e934cdf32261a804b184e06ae", "messageHeadline": "Add eager support for aten:: equal. (#12020)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 90, "deletions": 4}}, {"node": {"oid": "0ee0b8cf1853fcf1fb3e2906122474f43fe548dd", "messageHeadline": "Disable sequence-type tests since C# infra doesn't support well (#12037)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "da133ad3d8b6873cc400d4065aa4f524ba15d021", "messageHeadline": "Add FastGelu to kernel explorer for profiling. (#11995)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 295, "deletions": 72}}, {"node": {"oid": "fdf12a5c3516a7a28fb2f40a446a65b5e7f0acc5", "messageHeadline": "Fix windows eager build break by pinning to torch version 1.11.0 (#12\u2026", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "04f7c2deda004a9df09647edadf7d301ac2b4e24", "messageHeadline": "FP16_Optimizer Support for more Deepspeed Versions (#12046)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 21, "deletions": 26}}, {"node": {"oid": "ecca6f4d164b08e1e1b12aa1cd6f8c7557b0654b", "messageHeadline": "Move beamsearch shared initializers from subgraphs to main graph (#12\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 300, "deletions": 87}}, {"node": {"oid": "9f260fb60f1f6658e0684df801d069edac6af26b", "messageHeadline": "resolve comments", "author": {"name": "zhijxu", "email": "zhijxu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "100aebbd26f272b360e6853565043759cb473fd0", "messageHeadline": "resolve comments", "author": {"name": "zhijxu", "email": "zhijxu", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2295b24cd56917f34f153adc1f08754f5c914370", "messageHeadline": "support optimizer opt for deepspeed 0.5.9", "author": {"name": "zhijxu", "email": "zhijxu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "102d01b206247ae53634d1a6868e1749d4397898", "messageHeadline": "update roialign cuda impl to onnx opset16 (#12036)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "0ce14c70689d0b38a61991ac75912f1e927cd8c5", "messageHeadline": "Fix windows cpu build VS2022 (#12032)", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 34, "deletions": 28}}, {"node": {"oid": "6e8edfff0c5d6bfeeeba087a4018fb390d68c42e", "messageHeadline": "Separate training apis from shared core apis (#12027)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 162, "deletions": 126}}, {"node": {"oid": "c8cd36da018a117a042f194bff4720c9d39850bd", "messageHeadline": "Resize optimization for all architectures (#11956)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 2, "deletions": 14}}, {"node": {"oid": "4eb54ff9a5b38532e27cf43047e9e0e8bf6331c6", "messageHeadline": "Add warning about future computation change for ConvTranspose with au\u2026", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "8ba8146650222e96b607d96793b45410119df106", "messageHeadline": "[TVM] handshake mechanism for support of TVMso EP (#11437)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 217, "deletions": 6}}, {"node": {"oid": "c0dd9be7ba18c19bec5dbc90d3c9e2ec47f24c3d", "messageHeadline": "Bump electron from 13.6.6 to 15.5.5 in /js/web (#11884)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "0702364d7ae7b169670f2ffecc1bc5f13ab15353", "messageHeadline": "[js/web][bugfix] fix negative axes for unsqueeze (#11944)", "author": {"name": "Yosshi999", "email": "Yosshi999@users.noreply.github.com", "user": {"login": "Yosshi999"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9be2b6046bfb299b87e30502abda6ed4c003cf04", "messageHeadline": "convert_beam_search supports large gpt2 model (#11989)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 240, "deletions": 153}}, {"node": {"oid": "4552dd38c60e7c3e7a954fb53e08882bcfab1425", "messageHeadline": "[DML EP] Pad operator: Handle negative pad counts (#11974)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 19, "deletions": 1}}, {"node": {"oid": "d5fcb432fafc0ee37b466494c81bde49dd8ad07f", "messageHeadline": "Generalize native op creation (#11539)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 434, "deletions": 150}}, {"node": {"oid": "fc0143fe68390748e1001eb252d263c740eb2946", "messageHeadline": "DML EP ResNet50 opset 15 fails in ONNX checker for FusedBatchNormaliz\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "f0459943892f95d93a51635f9cdaefda6f7f2894", "messageHeadline": "[NNAPI EP] Update NNAPI headers (#11954)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 867, "deletions": 231}}, {"node": {"oid": "466b2d9f3daeb35132a637f5cc8420636e636ed2", "messageHeadline": "[C# Tests] Add support for double tensor output in TestPreTrainedMode\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 25, "deletions": 6}}, {"node": {"oid": "7d712c8f8bbf1ded56d408bc3da625173368d42d", "messageHeadline": "Fix WinML Tests are still targetting deprecated (deleted) experimenta\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 23, "deletions": 31}}, {"node": {"oid": "bd973bcf1ebe5b3622d345d0d17471ff39cff64c", "messageHeadline": "[js/rn] upgrade dependencies for e2e test (#11863)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1061, "deletions": 994}}, {"node": {"oid": "8cd02508c84ddaf62ec0b619fe2f86bbe453c67a", "messageHeadline": "Include opset 15 in Conv+BatchNormalization fusion (#11960)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "68afa2d362f577b9451e688539297415ac1c3bb5", "messageHeadline": "Bump async from 2.6.3 to 2.6.4 in /js/react_native/e2e (#11280)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "9583841ef79af0b361ac0562770df0dc2a5af1f1", "messageHeadline": "Improve performance of BiasGelu on oneDNN execution provider (#11935)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 22, "deletions": 23}}, {"node": {"oid": "f72288b453bedfc5ae2d1eab4725c014862db8d1", "messageHeadline": "Fix a couple of typos (#11943)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "dc5d6b9515b0a268bccc92e86601fd165c3de413", "messageHeadline": "register signal ops for opset 17 (#11778)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 1075, "deletions": 1980}}, {"node": {"oid": "d25cf4df263e458e89d49bcb4b05efa8d5fe48e7", "messageHeadline": "Merge branch 'master' into training_dev/on_device_poc", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9423, "deletions": 3815}}, {"node": {"oid": "f4ba199baddbf176369a8fba7ca53a328a361a6c", "messageHeadline": "Optimize FastGelu with float2 and float4 vectorized kernels on ROCm (\u2026", "author": {"name": "Hubert Lu", "email": "55214931+hubertlu-tw@users.noreply.github.com", "user": {"login": "hubertlu-tw"}}, "additions": 321, "deletions": 144}}, {"node": {"oid": "088bc7494bf7feb3b1aa9a8460f1ba0d5123f4a9", "messageHeadline": "Deprecate APIs returning raw ptrs and provide replacements (#11922)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 346, "deletions": 83}}, {"node": {"oid": "b1411c83578a2464647d66ad9b0ddde1c0f0a6cd", "messageHeadline": "Restructure function inliner (#11731)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 426, "deletions": 416}}, {"node": {"oid": "c2fd5ccbe94bc3e62eca770b17b0b631883419a3", "messageHeadline": "Redesign InPlaceAccumulator op (#11842)", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 300, "deletions": 73}}, {"node": {"oid": "0d6cbc6e57c2ecd0722cefaa87d300f301864e3b", "messageHeadline": "fix memory profile for partial graph run (#11911)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 225, "deletions": 141}}, {"node": {"oid": "fa7f80c84739223ddc965fbfc6d232737e76b284", "messageHeadline": "Eager mode: Argmax and fixup max and min. (#11861)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 119, "deletions": 48}}, {"node": {"oid": "2c4e4b6afc396b1de29b34727edb151c58ee0d89", "messageHeadline": "MT5 onnx conversion for beam search (#11958)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 234, "deletions": 125}}, {"node": {"oid": "607b7df060c2f4e83bd79647cdf7fe742da59749", "messageHeadline": "Allow saving on CPU usage for infrequent inference requests by reduci\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 654, "deletions": 569}}, {"node": {"oid": "c398ad513f4a6e6577af6d33f0601ae180862acc", "messageHeadline": "Fix orttraining-linux-ci-pipeline - Symbolic shape infer (#11965)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "17a8ecee6fb4bc8fafc7b45bee363cebb28eb5df", "messageHeadline": "fix win build errors (on device training) (#11844)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 78, "deletions": 58}}, {"node": {"oid": "e24349b8f26645a5069ac4477b1f72b60d079661", "messageHeadline": "Optimize t5 encoder in beam search (#11926)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 335, "deletions": 67}}, {"node": {"oid": "f6d2fe831101c1d2f6fb2b6959a1081b8c6361ef", "messageHeadline": "MeanVarianceNormalization CPU EP axes attribute validation (#11925)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "f54476a42f3ccb2af11197cfe8a03ee3107512e4", "messageHeadline": "Dll version fix ovep4.1 (#11953)", "author": {"name": "Preetha Veeramalai", "email": "preetha.veeramalai@intel.com", "user": {"login": "preetha-intel"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "fac8dae9df8676c5e88f83e87269c4e62d386065", "messageHeadline": "Add support for gradient clipping, AdamWOptimizer and tensorseq as in\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 743, "deletions": 675}}, {"node": {"oid": "f14f0e19eceaa32ea179829c5314676df8011a51", "messageHeadline": "Fix trainer (#11933)", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "2229c485478e323a3ac2cac46079c2df6518ae38", "messageHeadline": "fix mpi in training build  (#11855)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 22, "deletions": 15}}, {"node": {"oid": "03beed0ceb691e8b91df5fb266385488358d07e6", "messageHeadline": "Remove Cast before and after Gelu (#11885)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 140, "deletions": 4}}, {"node": {"oid": "a36e92d86edcdbc04c471fe4ea83b04629540b6c", "messageHeadline": "Offline tooling readme (#11920)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 76, "deletions": 0}}, {"node": {"oid": "4bf22e2a40b5cbc8193f985bd784693d07e0e3b5", "messageHeadline": "Update ONNX to 1.12 (#11924)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 1089, "deletions": 148}}, {"node": {"oid": "64f95d400a788cf165adbf23dc84079457c00b15", "messageHeadline": "Update DML 1.9 Nuget package to fix WindowsAI nuget pipeline build is\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "3b1224dc0838e23c9933d4dfb64568898c7ffa8b", "messageHeadline": "Add .net6 support to the C# nuget package. (#11908)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 530, "deletions": 339}}, {"node": {"oid": "8c8a781cdb7a55d5cb9a5003f0cf4dcac780c6fe", "messageHeadline": "fix: handle setBindingDimensions return value in TensorRT EP (#11929)", "author": {"name": "Arseny", "email": "82811840+senysenyseny16@users.noreply.github.com", "user": {"login": "senysenyseny16"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "5646410f659f791ce1a9dd60a66e17f5f284db19", "messageHeadline": "Enable Pad test cases with initializer inputs only when building NNAP\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "61a74f2f4d65dbaa6e20b17a11ec66a5d55c5e7f", "messageHeadline": "Mohsin/enable dynamic shapes (#11867)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 138, "deletions": 31}}, {"node": {"oid": "b20daeda81d008eb0a76485ab233c1a1e4c4e762", "messageHeadline": "Update Linux Multi GPU TensorRT pipeline to TensorRT 8.4 (#11923)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 17, "deletions": 10}}, {"node": {"oid": "859ef277a0d75e90bdfc0b3f35d4f9194791aabc", "messageHeadline": "apply zcode changes to the beam search op (#11880)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 268, "deletions": 82}}, {"node": {"oid": "cefceff5c91b5cf5fda076027dbf29942585b960", "messageHeadline": "Mark the end of APIs for release 1.12 (#11914)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "ca35ea417a672faa39971d351dd3e5cf11e0a5e9", "messageHeadline": "[EP-Perf] Install new wheel>=0.35.1 dependency (#11917)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "7f1e9e8c67b2d88a13c48d8e785de9c5dd655cee", "messageHeadline": "Bash: there should be a whitespace after not operator. (#11910)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "457ce6cb8910bdfa7e0117030e6c83838b946ea6", "messageHeadline": "Make symbolic shape inference script support external weight (#11909)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 35, "deletions": 1}}, {"node": {"oid": "c1577d08ca22492e32631faa092f9f2608923131", "messageHeadline": "DML EP QuantizeLinear defer axis validation for test_quantizelinear_c\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 10, "deletions": 7}}, {"node": {"oid": "0736c604c7c84199206e3743039d09d43d2286fb", "messageHeadline": "Fix building DNNL EP with GCC 12 (#11667)", "author": {"name": "Chih-Hsuan Yen", "email": "yan12125@gmail.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "267a424e5295633a9ad61058b8ff96400b9b0489", "messageHeadline": "Retry Rework execution frame to reduce memory allocations (#11897)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 323, "deletions": 228}}, {"node": {"oid": "6ee2c1b5fce1e1a2b0185835457316e3c2065175", "messageHeadline": "Remove temperature input from BeamSearch operator (#11896)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 13, "deletions": 55}}, {"node": {"oid": "eb41bfb7b585e8eed3e876f0fd2b085ee2879fc9", "messageHeadline": "Fix graph viewer to proto (#11862)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 22, "deletions": 1}}, {"node": {"oid": "52f2b3bf89d525d1f8ecbd099c8299004f4ceef0", "messageHeadline": "[DML EP] Remove suffix removal adhoc logic for fusedNodeArgNames (#11\u2026", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 57, "deletions": 81}}, {"node": {"oid": "a3ec2d6d15161c0b726f67117aeb654b52368218", "messageHeadline": "changes to c api + bug fix (#11858)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 28, "deletions": 44}}, {"node": {"oid": "f97bd38c4fe3b745d67f5fa05fa4b10f258fbfab", "messageHeadline": "UEP 4.1 release (#11834)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 276, "deletions": 10}}, {"node": {"oid": "a93fe7824a72f739e7feb79ec2b670c947479d78", "messageHeadline": "Update EP compile API deprecation warning message. (#11808)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "f70201c8017bd064f1da13ceb3d3da76846c387d", "messageHeadline": "Make sure the command works in both centos and ubuntu. (#11894)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "149412042397b19a1c916ba37bf3acdfa3933e51", "messageHeadline": "[NNAPI EP] Unsqueeze op support (#11864)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 115, "deletions": 28}}, {"node": {"oid": "4ac72e305c5fe61abf6f3f4ba712e9198335ab85", "messageHeadline": "NHWC Resize optimization (#11825)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 1211, "deletions": 467}}, {"node": {"oid": "adcf7e66c893870f3810a18b1dcad7760cbcc5b9", "messageHeadline": "[NNAPI EP] Pad Op (#11860)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 304, "deletions": 121}}, {"node": {"oid": "ad4abbd75e7beac722a7416b110b75bb290caefc", "messageHeadline": "[EP-Perf-Dashboard] Add support for TensorRT 8.4 to EP Perf Dashboard\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 68, "deletions": 10}}, {"node": {"oid": "8bb0062873ddfef7f5938bfae3f122d76c112918", "messageHeadline": "add manylinux_2_27 CPU wheel (#11886)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 362, "deletions": 26}}, {"node": {"oid": "d2cbae3a04dc23a39c008122c57577fed776f849", "messageHeadline": "Revert \"Refactor ExecutionFrame and SessionState to reduce memory all\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 232, "deletions": 326}}, {"node": {"oid": "bd65acd08d06f430c73cbc1372f2aea13141130f", "messageHeadline": "Share execution context memory between TensorRT subgraphs (#11859)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 100, "deletions": 18}}, {"node": {"oid": "10478a09ca101a8cda92aa514e556bbeb6fb112f", "messageHeadline": "Revert \"add manylinux_2_27 wheel (#11832)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 352}}, {"node": {"oid": "2ecba6fd25cce35c03f6b3891753e1bb1ee6275c", "messageHeadline": "Refactor ExecutionFrame and SessionState to reduce memory allocations\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 326, "deletions": 232}}, {"node": {"oid": "3d99f16e98822ffc9e295484ee7e519af22411a7", "messageHeadline": "Merge pull request #11827 from microsoft/user/dwayner/DmlEp1.9", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 6998, "deletions": 2044}}, {"node": {"oid": "df5ee6aa4ee24172fe21dc07c4a76ebc9ed6f54c", "messageHeadline": "[TensorRT EP] support TensorRT 8.4 (#11866)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 277, "deletions": 44}}, {"node": {"oid": "fe7b8b80ae35c62e5bf91eed068be4279b76173d", "messageHeadline": "Revert BatchNormalization change for now, falling back to CPU on mixe\u2026", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 99, "deletions": 7}}, {"node": {"oid": "babd6e3fcdcbf5246e675e8d4dcb3e07e581e176", "messageHeadline": "Update DirectML preview package with unmangled names", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 9, "deletions": 58}}, {"node": {"oid": "3f8c9146d5874e44986ef43de48d870da212ba0b", "messageHeadline": "ppc64le: specialize generic 'mlas' functions to use VSX instructions \u2026", "author": {"name": "Maxiwell S. Garcia", "email": "maxiwell@linux.ibm.com", "user": {"login": "maxiwell"}}, "additions": 23, "deletions": 2}}, {"node": {"oid": "d64f23fec0316b4e298bc8c6fbf285e2e73f2485", "messageHeadline": "EP factory creation cleanup and enhancements. (#11798)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1067, "deletions": 558}}, {"node": {"oid": "f63e28c92fcb5b610a2eb11cc2a32f27e5b19e66", "messageHeadline": "C API version 0.001 (#11758)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 666, "deletions": 137}}, {"node": {"oid": "1a1c360a804fba35fea25e4c9f27244848e6a4fb", "messageHeadline": "[NNAPI EP] Add Gather op support (#11824)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 364, "deletions": 107}}, {"node": {"oid": "02457ec30a3b0270c9d38c5581e6b6b982c17233", "messageHeadline": "[CUDA] GatherElements[Grad]/ScatterElements Bugfix and Perf Improve (\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 996, "deletions": 1460}}, {"node": {"oid": "a805a49363021a6f75f12130bb32e48a909a0ff2", "messageHeadline": "Move OrtValueVector from onnxruntime-training to onnxruntime (#11176)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 189, "deletions": 171}}, {"node": {"oid": "ff8b173286248480b7b316b4c97b0031a4190b11", "messageHeadline": "Typo in DirectML.Debug.dll", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "508c76a246a2cbebcd2a54b2d4ff371413d8ab02", "messageHeadline": "Add missing DirectML.Debug.dll", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 25, "deletions": 12}}, {"node": {"oid": "e3ec30efb65d596a6aab5dea4b8edbdb8c1a1370", "messageHeadline": "Add missing GELU to ApiHelpers.h", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "4c1a410d54bc63a0d18df940af45260018f65b94", "messageHeadline": "Unmangle DML preview package filenames", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 56, "deletions": 5}}, {"node": {"oid": "bbace23d0cc2286ecd7600f1cc5427bbccfcfddd", "messageHeadline": "add manylinux_2_27 wheel (#11832)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 352, "deletions": 15}}, {"node": {"oid": "51ed27cf22230125c638f51edeb7f6a46bddcf95", "messageHeadline": "Delete win-gpu-cuda-10-2-pipeline.yml (#11847)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 127}}, {"node": {"oid": "3cbbf9dcaed4db1d7d2c8014223514e6c06287c2", "messageHeadline": "Fix wasm static lib in sub-project (#11671)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "e8b0d2407157216a898d0575cf6fe03df6ece389", "messageHeadline": "Support per-test tolerances for ONNX tests (#11775)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 253, "deletions": 106}}, {"node": {"oid": "d936751aada2d1d66c902a8c37af413aba7d18a2", "messageHeadline": "QlinearConv threading adjustments (#11228)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 210, "deletions": 52}}, {"node": {"oid": "80d8c4c7ff4075495087ecde08892113940a5d24", "messageHeadline": "add data type check before quantizing (#11840)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 92, "deletions": 45}}, {"node": {"oid": "607afbe1c02cabc719cdb449ea6a363d9a30fafb", "messageHeadline": "fix valgrind warnings:Conditional jump or move depends on uninitialis\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "52f6db19da584d7e41c869ab8c7c8197db1e08ae", "messageHeadline": "Python backend: use packaging.version to parse ONNX version (#11800)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "f6d2b629a04dbb135e2aa7a0a3f86afeab7df102", "messageHeadline": "Add kernel explorer (#11779)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 387, "deletions": 0}}, {"node": {"oid": "6bf6bac1fdd4a386df1cef8dcd75c5673875dc17", "messageHeadline": "Add patching of xnnpack CMakeLists.txt to allow building with Emscrip\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 74, "deletions": 56}}, {"node": {"oid": "63c483a998cf6116c23837bba8c00dbdcedd8234", "messageHeadline": "1.12.0 is the right TBD instead of released 1.11.0 (#11817)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "aef53e2b0d168d714cb49ab48db221f7edb5b62c", "messageHeadline": "Support uploading EP perf data to a configurable database. (#11819)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 7, "deletions": 8}}, {"node": {"oid": "a93ebd2503fc364f8ada6fa8ce32026c39e5d0a2", "messageHeadline": "Move tvm pipeline to Github Actions (#11721)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 43, "deletions": 53}}, {"node": {"oid": "b0e027c661769f4d5edd5a893e4cb7bd6b8ff053", "messageHeadline": "Add aten::_softmax to eager ops. (#11820)", "author": {"name": "Wil Brady", "email": "25513670+WilBrady@users.noreply.github.com", "user": {"login": "WilBrady"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "7582644f579fbbc584da9e8dcad03b8eb7204fc1", "messageHeadline": "cmake changes for SNPE EP (#11821)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 48, "deletions": 38}}, {"node": {"oid": "04dd6639de1d2dcb8ff410994ebc68000ab448a9", "messageHeadline": "And appease the time wasting formatting tool now -_-...", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "2bc487a816b55445d061c2e573c5f52fab5917dc", "messageHeadline": "Appease flaky flake tool", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "50e0a193c8dfbc512564616a594e23a7812cf183", "messageHeadline": "Merge branch 'master' into user/dwayner/DmlEp1.9", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 13404, "deletions": 5320}}, {"node": {"oid": "76024b8a6a4c45a5453df8e26fda91d190b13d9a", "messageHeadline": "Update DirectML.dll to 1.9.0 Preview", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "0869f4f4ea9abeb4edf2e5d5b570880f77f81bfa", "messageHeadline": "ppc64le: optimizing the MlasRequantizeOutput() with VSX (#11659)", "author": {"name": "Maxiwell S. Garcia", "email": "maxiwell@linux.ibm.com", "user": {"login": "maxiwell"}}, "additions": 192, "deletions": 0}}, {"node": {"oid": "fb88efbe1813df510ff5c9160a09a880ce68168f", "messageHeadline": "End to end run pass (on device training) (#11694)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 820, "deletions": 340}}, {"node": {"oid": "def78a1b815ae2b60d19df97c352230591e7d7a6", "messageHeadline": "Support T5 in BeamSearch operator (#11450)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3705, "deletions": 1462}}, {"node": {"oid": "c1b5f343621bdc63440084b7054992ead0510a5a", "messageHeadline": "DML EP BatchNormalization-15 (#11814)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 6169, "deletions": 1855}}, {"node": {"oid": "768b9cfb60cc981a823144d548597d0df47afd46", "messageHeadline": "Fix GetDirNameFromFilePath to support forward slash in windows (#11793)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 27, "deletions": 3}}, {"node": {"oid": "a61c38e4f42cf57860394c0a2548e9e6b0266deb", "messageHeadline": "Add ability to author float initializers (#11752)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 100, "deletions": 67}}, {"node": {"oid": "5562b47f0687ae019c9ae66bf4692e2a44b0e69d", "messageHeadline": "missing #include <thrust/count.h> in non_max_suppression_impl.cu (#11\u2026", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d4ea59654c83cb6f7e578c6d6c7270a57ec990b2", "messageHeadline": "make xnnpack build for ort-web (#11745)", "author": {"name": "Guenther Schmuelling", "email": "guschmue@microsoft.com", "user": {"login": "guschmue"}}, "additions": 12, "deletions": 3}}, {"node": {"oid": "f745eb1d3f6f1bcfaddd2717d8a491f9339a04e7", "messageHeadline": "fix gradient ut (#11797)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 82, "deletions": 44}}, {"node": {"oid": "5ecfaef042380995fb15587ccf6ff77f9d3a01d2", "messageHeadline": "ATen Fallback for Inference (#11597)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 755, "deletions": 267}}, {"node": {"oid": "927bac0f862d56bed135c451424ae39467e5b530", "messageHeadline": "Rework allocator sharing to work for multiple devices. (#11700)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 762, "deletions": 538}}, {"node": {"oid": "5e54611427d6c98cea014c200b24edee57c92295", "messageHeadline": "DML EP add Trilu-14 and Resize-13 nearest mode and others (#11782)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 311, "deletions": 57}}, {"node": {"oid": "0f0b640b4b1a142a17574fa60159d33c34c2e614", "messageHeadline": "Reformat build.py for WindowsAI branch (#11794)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "8156b9370c635530a2080850537c0b8c01a2323c", "messageHeadline": "[Abseil] Adding URL_HASH so that an existing archive can be used from\u2026", "author": {"name": "Alex Fuller", "email": "boberfly@gmail.com", "user": {"login": "boberfly"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "540935aacea6b6ab70a095cae61b14d3b87d96b1", "messageHeadline": "lr scheduler implementation (on device training) (#11714)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 4019, "deletions": 87}}, {"node": {"oid": "913100885b2f973ddf749a9d9d1a0b85fbf72db3", "messageHeadline": "Remove the redundant black check in CI (#11790)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 2, "deletions": 9}}, {"node": {"oid": "79db92f8fea2f14b354ea4bad8d5b17f19983efb", "messageHeadline": "clang-format signal_defs.cc (#11767)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 259, "deletions": 263}}, {"node": {"oid": "f5fe4f253c336a77e575a090a11bd180ff087d63", "messageHeadline": "Registered Softmax/Hardmax/LogSoftmax-13 as Versioned Operator (#11787)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "750cb42f87b5c1328fb2f7959e8930d1f7f3ae07", "messageHeadline": "Bump protobufjs from 6.10.2 to 6.11.3 in /js/node (#11722)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 8, "deletions": 22}}, {"node": {"oid": "bc4c7710786e1e3ec4ebd310983b14758f45f621", "messageHeadline": "Bump protobufjs from 6.10.2 to 6.11.3 in /js/web (#11723)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 10, "deletions": 24}}, {"node": {"oid": "eeeb249a279ef6fa106429e1c8654d2423591769", "messageHeadline": "Update onnxruntime_providers.cmake to remove the reference of \"onnxru\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "331c387f4a5862c033492635349e8701084d1e34", "messageHeadline": "[TVM EP][DOC] Documentation update for TVM EP due to the addition of \u2026", "author": {"name": "Alexey Gladyshev", "email": "wotpricol@mail.ru", "user": {"login": "KJlaccHoeUM9l"}}, "additions": 253, "deletions": 74}}, {"node": {"oid": "7f8d0ba824ec3e38cd42d59b75ffcd2ddf7ad48e", "messageHeadline": "Update comments in Android workflow (#11311)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f6f457aa5734e095726bdec1e46eb6f427c5fb1d", "messageHeadline": "not remove relu/clip for symmetric activation (#11696)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 352, "deletions": 90}}, {"node": {"oid": "aa3a825816555d1f859385a256804dfceefebd2f", "messageHeadline": "Added Softmax/Hardmax/LogSoftmax-13 (#11772)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 189, "deletions": 17}}, {"node": {"oid": "40f4304c7d567436e39006cdf9242f771858bc8e", "messageHeadline": "[Fix #11447] Use correct type for tensor shape vectors (#11448)", "author": {"name": "GPhilo", "email": "4441724+GPhilo@users.noreply.github.com", "user": {"login": "GPhilo"}}, "additions": 12, "deletions": 11}}, {"node": {"oid": "b4f1e769c09f75af51623c2e296017c491adffe5", "messageHeadline": "Add Mac Silicon/M1 Wheel (#11591)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 88, "deletions": 1}}, {"node": {"oid": "40d2c98e4d64d7931fcbb682c81dc53988c48dcc", "messageHeadline": "[js/web] fix ORT Web dependency version mismatch", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "8fb38e8a548009eb5301bb5160acd95b6ed10a05", "messageHeadline": "fix cmake warning (#11742)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 18, "deletions": 20}}, {"node": {"oid": "9e33bfd29bfda14ad1844bc8279c7c3e37311cc0", "messageHeadline": "Bump simple-plist from 1.3.0 to 1.3.1 in /js/react_native/e2e (#11712)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "908e19dc1688b7a04b34e885747225c6dcfd7502", "messageHeadline": "[FIX] using torch.version.cuda/hip to ensure build ORTModule Torch C+\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "981d45d8d54b0c085a9e424366c99b42f4514c67", "messageHeadline": "Add binary comparators to the OneDNN (dnnl) execution provider (#11641)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "4296968f2057743d43ebf6f3aebd48571fe3f9d7", "messageHeadline": "[TVM EP] update set input method for VirtualMachine (#11674)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 20, "deletions": 8}}, {"node": {"oid": "d5e34acb826e3e65b966a75e144987d57ec29290", "messageHeadline": "Remove git and python packages from the docker images used by Zip-Nug\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 247, "deletions": 146}}, {"node": {"oid": "3c1dd9514d95985b954b83a97c1bb4492e851b63", "messageHeadline": "Revert \"fixed point based requantization on arm64 (#11540)\" (#11732)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1123, "deletions": 3796}}, {"node": {"oid": "ef64b2ee5200f15e2c51039753f1b7907cde24d7", "messageHeadline": "Fix clash between QDQ propagation and TransposeOptimizer (#11636)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 136, "deletions": 31}}, {"node": {"oid": "95a16c1ffe93c061c70e2d4ae4b0950bf40ae2d0", "messageHeadline": "Snpe ep (#11665)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 1925, "deletions": 53}}, {"node": {"oid": "98960c53febe36a734465d1f6954104151c34992", "messageHeadline": "Replace ORT's function shape inference with ONNX's (#11538)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 138, "deletions": 128}}, {"node": {"oid": "ec05313cd9623dc2f644181247273fb13c503e60", "messageHeadline": "Split the GPU pipeline to 3 different machine pools (#11724)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "4445dd6bc14ad6d145fa2321b4c61c84f7cd9100", "messageHeadline": "XNNPACK EP (#11445)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2158, "deletions": 257}}, {"node": {"oid": "ce4ac6d3280f61b8c30801d044f32d423f69559c", "messageHeadline": "Optimizer - add missing supported version for BiasSoftmaxFusion (#11616)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 88, "deletions": 15}}, {"node": {"oid": "196cd7aed196f74dbc80062c168ec531a4375c10", "messageHeadline": "Pylint fix after #11647 (#11704)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 77, "deletions": 30}}, {"node": {"oid": "1f2c92673b2edf75460478d406f84aead7a140f3", "messageHeadline": "fixed point based requantization on arm64 (#11540)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 3794, "deletions": 1120}}, {"node": {"oid": "453c57f92f5294417f69d81a240484b7d59938f2", "messageHeadline": "Revert \"Directly use memory mapped data for external data initializer\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 98}}, {"node": {"oid": "54d1573d2f6ab27cdf6e1bf5848a51bc03e0853c", "messageHeadline": "[ORTModule] Enable SimplifiedLayerNormalization Fusion (#11580)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 10, "deletions": 76}}, {"node": {"oid": "03abcb06406f1014ab902833ccb91b402589e25d", "messageHeadline": "Correctly unpack tensor values (#11639)", "author": {"name": "Chih-Hsuan Yen", "email": "yan12125@gmail.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2ac3649752a97ace874bd2ce57dae2325e43c361", "messageHeadline": "Update requirements.txt (#11682)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "f437945926ae121ff6892a3c235cb61240fc1121", "messageHeadline": "fix output shape of ReduceMin/ReduceMax in calibration tool (#11647)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 150, "deletions": 156}}, {"node": {"oid": "004560f1fe7044f36837de61838505503518bb9e", "messageHeadline": "[js/rn] upgrade dependency packages' version (#11586)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1480, "deletions": 1489}}, {"node": {"oid": "74bc4c07f6997eeaa8a063d7629ef47eef678073", "messageHeadline": "Fix C# and numbering (#11643)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "1c316d0e3954b2e50ef3e6f6ece20ea26f2fdf44", "messageHeadline": "Parameter,Module and Optimizer changes (#11494)", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 692, "deletions": 147}}, {"node": {"oid": "3076061acae9685102ccefb33f24313c99b32cb3", "messageHeadline": "Fix a comment in test_op_reshape.py (#11606)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "22739137c419bb8a86b8f9ff5c8def60b63370c7", "messageHeadline": "Update signal op defs to match onnx17 defs, and add more tests (#11631)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 665, "deletions": 272}}, {"node": {"oid": "e1c63cb06a3ef15f18a08e8543f30134550f9d61", "messageHeadline": "Merge branch 'master' of https://github.com/microsoft/onnxruntime int\u2026", "author": {"name": "pengwa@microsoft.com", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 73887, "deletions": 52090}}, {"node": {"oid": "c318b19307800621c14304f1a8f0ad20b59e399d", "messageHeadline": "Add support for BCEWithLogitsLoss (#11630)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 462, "deletions": 35}}, {"node": {"oid": "4fabc400dedd41b8fb9f5805ea570973cf9f3134", "messageHeadline": "Fix CUDA 11.6 build error on Windows (#11578)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "7e6d052275b0d0ad377cda95cf0d7e6e819816be", "messageHeadline": "Add better error message for subgraph output coming directly from out\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 47, "deletions": 32}}, {"node": {"oid": "a7fa73528605dd878a5c88cb159454c3612086f2", "messageHeadline": "Merge remote-tracking branch 'origin/master' into WindowsAI", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 11271, "deletions": 10100}}, {"node": {"oid": "b67c0f639ce5c9b231c275d08539f0c50c0ced51", "messageHeadline": "Remove filter_mode input from pyflakes GitHub action (#11644)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "44f7b1bf2ced2331d81e3801922881acd899d3b0", "messageHeadline": "MTA AdamWOptimizer (#11506)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 4519, "deletions": 48}}, {"node": {"oid": "02724c54ff15aa974de63841527aa6b57dd3e7de", "messageHeadline": "[CUDA] Implement BitmaskDropout, BitmaskBiasDropout and BitmaskDropou\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1674, "deletions": 618}}, {"node": {"oid": "eadb1a3128447cabcf16eebb6e4519ebe1e94376", "messageHeadline": "Speed Up GradientChecker Running (#11579)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 739, "deletions": 917}}, {"node": {"oid": "6a45f9f05967703ac26cd52dae7443fa8937f97c", "messageHeadline": "Pin protobuf version to 3.18.1 (#11645)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "006597b9b8d581c4db5e4294863a33c65cb49fb5", "messageHeadline": "Microsoft mandatory file (#11619)", "author": {"name": "microsoft-github-policy-service[bot]", "email": "77245923+microsoft-github-policy-service[bot]@users.noreply.github.com", "user": {"login": "microsoft-github-policy-service[bot]"}}, "additions": 41, "deletions": 0}}, {"node": {"oid": "f0dff6bb74fbc2a37f8c08e2615308d4f5ed26a7", "messageHeadline": "[js/rn] add expo config plugin support (#11556)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 254, "deletions": 5}}, {"node": {"oid": "d03d7afef847b4ce20f0f8918e4634941ea9827f", "messageHeadline": "Fix build errors when building with enable_memory_profile (#11617)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "6e65bac5c2efec68491d109b7ddc953169684a81", "messageHeadline": "Memory usage optimization in LongFormer Attention (#11611)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "883e4bc3417577b631a5f292e9074d5c0e4dff5c", "messageHeadline": "Update the 'Linux-GPU-EP-Perf' pipeline to build ORT from source by d\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "427230431a88a016b3f35893bece52878941a588", "messageHeadline": "Fix torch cpp ext build when CPU wheel is installed but GPU card is p\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 25, "deletions": 13}}, {"node": {"oid": "147a1737f932a093cc05bb029c86c3f5b0add592", "messageHeadline": "MatMul postop fusion for dnnl ep (#11565)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 1616, "deletions": 151}}, {"node": {"oid": "4e9ad7b6ae308e506f530cd52af92664a719c3b3", "messageHeadline": "Update .flake8 to exclude .git directory (#11615)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "3a22a866a164a0d21499af78ca9fb94ce79af78e", "messageHeadline": "On device training offline tooling (#11520)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1434, "deletions": 2}}, {"node": {"oid": "e3a2d5cca8bcefe064f83d57e46ea51ddb2b16e8", "messageHeadline": "Add additional python requirements (#11522)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "69aaf03345ec2acf40c9deaa5170d15652deb0dc", "messageHeadline": "allow catch all exceptions (#11498)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a67994316a856e6dd0b1619b18467073e75b3a00", "messageHeadline": "Update rocm ci to ROCm5.1.1 + torch1.10.0", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 88, "deletions": 766}}, {"node": {"oid": "abecb56832ac7cadf3466a4bb8078dc1e30094e8", "messageHeadline": "fix buid break (#11492)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d2519ec0c26c31e632e4dd72ad9550f1f9a18df7", "messageHeadline": "DirectML EP add CastLike-15 and Reshape-14 (#11568)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 68, "deletions": 41}}, {"node": {"oid": "436c4f9b7997088ae1f0dc31654b980630b870a1", "messageHeadline": "Add BFloat16 (bf16) support for ATen (#11546)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 22, "deletions": 3}}, {"node": {"oid": "3a867d83d5c8af4d308b7e4212f99d96613a7ae0", "messageHeadline": "DirectML opset14 type updates for Add/Sub/Mul/Div and Relu/PRelu (#11\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 71, "deletions": 102}}, {"node": {"oid": "e45197fa8cc32310ca4e994d1fdb7304a5bc94b0", "messageHeadline": "[trt-ep-perf] Fix upload time of EP perf data (#11531)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 193, "deletions": 52}}, {"node": {"oid": "8092d9f9a2a7dbe3bd2f66734ab105847d63a3c9", "messageHeadline": "[TVM EP] Support inference by shared library created by TVM (#11389)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 609, "deletions": 59}}, {"node": {"oid": "48efeca66cbc7eada72a0c59a87476a75b36aa14", "messageHeadline": "[trt-ep-perf] Fix bug that suppresses latency gain reporting (#11321)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 51, "deletions": 16}}, {"node": {"oid": "782f9e394df0d88591a856b31f781e69deb46fa8", "messageHeadline": "[CoreML EP] Fix condition in PRelu op supported check. (#11543)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "deef21477232635a33d9e7a0d926c01e5ba5a77c", "messageHeadline": "Update gather to use multiple threads (#11524)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 73, "deletions": 90}}, {"node": {"oid": "5eaa8939361d1dab40a191ced6acb7d13d09a025", "messageHeadline": "[CoreML EP] Add support for PRelu (#11474)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 185, "deletions": 22}}, {"node": {"oid": "d9c9adb78bf2a1ae4e272cb3888301d08fd5acd2", "messageHeadline": "Add python static type checking in CI checks (#11518)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 27, "deletions": 8}}, {"node": {"oid": "c556f5f22fd255c2673590dcd6b7f875abc37aa6", "messageHeadline": "Add AMD python package ROCm5.1.1+torch1.11 (#11516)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 47, "deletions": 47}}, {"node": {"oid": "62551946593382489b1f3dc4187adc0838ed472b", "messageHeadline": "All LearningModelSessions created from a common LearningModelDevice s\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 422, "deletions": 98}}, {"node": {"oid": "5709ed2e164b5b0e39d739ac10335d7bb2da2ec1", "messageHeadline": "Fix shellcheck warning (#11489)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "b14c1fd479465dd1bdbe54648855a68881ff93b9", "messageHeadline": "POWER: Optimize MlasQLinearAddKernelHelper() (#11454)", "author": {"name": "RajalakshmiSR", "email": "58744256+RajalakshmiSR@users.noreply.github.com", "user": {"login": "RajalakshmiSR"}}, "additions": 198, "deletions": 0}}, {"node": {"oid": "09590f013a974607cec5fe757874b4cff5aa1498", "messageHeadline": "fix windows ci debug build break (#11495)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "4aef7e3aabd5cd99f4d8275a926ab5865e2bbf44", "messageHeadline": "[CoreML EP] Add DepthToSpace op support (#11468)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 113, "deletions": 13}}, {"node": {"oid": "a3f05da33875d4e8383c0b91d82ecd338842ca1e", "messageHeadline": "Revert \"[TVM EP] update set input to remove excess copying inside TVM\u2026", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 8, "deletions": 16}}, {"node": {"oid": "ece1274ffa5468aa1f5b7468f4f6f3a0e939f2a3", "messageHeadline": "revert safeint version (#11500)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f94b25933a2fabcd71bf345dd59e6a2642cc64fe", "messageHeadline": "ci(cpplint): Ignore runtime/references warnings (#11499)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "2660eb836427807d843efd563366450af54bda72", "messageHeadline": "DML EP: Gelu (#11483)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 47, "deletions": 3}}, {"node": {"oid": "6f85d3e5c81c919022ac4a77e5a051da8518b15d", "messageHeadline": "fix(onnx_export): Extract arg value from torch Value (#11471)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 19, "deletions": 0}}, {"node": {"oid": "f5473596fa0db4f16046e86dc85013583360551f", "messageHeadline": "Change longformer default kernel (#11470)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 309, "deletions": 208}}, {"node": {"oid": "48ae27d57873928884714fa37e8f88e9ea3d73c1", "messageHeadline": "Update protobuf-java to 3.20.1 (#10420)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "207ad7eef9f91f88312cc28b2528bdc6362eaadb", "messageHeadline": "Remove spdlog from cgmanifest.json", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 10}}, {"node": {"oid": "027fc1d3912b0739113039c6ade32436a121d65f", "messageHeadline": "Completely delete ORT server", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 6475}}, {"node": {"oid": "903743e82361e51701de47265db1a29742cb25a9", "messageHeadline": "Delete unused TRT docker files (#11486)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 439}}, {"node": {"oid": "205b61c5d84989e4f8a538e7db125aee69c23960", "messageHeadline": "Fix bad merge in build.py", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 0, "deletions": 4}}, {"node": {"oid": "f82946c4a0f16bc54e5c7502fce75e7401a290f7", "messageHeadline": "Merge branch 'master' into user/dwayner/WindowsRiTest2", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 65076, "deletions": 43299}}, {"node": {"oid": "0ac2e6e546cf3591b93fcb463ea9510fa4153360", "messageHeadline": "Update install-entrypoint.sh: add version lock for NCCL (#11475)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d8a1531c375f6b4ecd5c5f38b6e1d194b0f40d9c", "messageHeadline": "CKPT API Implementation (On Device Training) (#11261)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1840, "deletions": 358}}, {"node": {"oid": "3437967e63f1cae12613b187c0a71687d08b0042", "messageHeadline": "[js/rn] fix CI packaging for react native E2E test (#11463)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 19, "deletions": 36}}, {"node": {"oid": "738d9b153cae5e7f586036da47d5fc1e984a32c4", "messageHeadline": "Consolidate several types into onnxruntime::ArgType. (#11430)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 39, "deletions": 43}}, {"node": {"oid": "288892335e7499c4b235aea96fc097e3775435a1", "messageHeadline": "[NNAPI EP] Add support for DepthToSpace Op (#11354)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 127, "deletions": 2}}, {"node": {"oid": "3b16fb2000574b5113cfe6853bd4a32c5ea5bb7b", "messageHeadline": "Delete java-test-final-jar-step.yml (#8894)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 80}}, {"node": {"oid": "c54106324548b39e286b6e91ea8be81166c37145", "messageHeadline": "Format coding conventions documentation (#11405)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 38, "deletions": 38}}, {"node": {"oid": "c2de603c1004c4317f5b72d395a2e74d7e57ee1f", "messageHeadline": "Contrib ops for TRT plugin: Disentangled Attention Plugin (#11287)", "author": {"name": "symphonylyh", "email": "31998628+symphonylyh@users.noreply.github.com", "user": {"login": "symphonylyh"}}, "additions": 43, "deletions": 2}}, {"node": {"oid": "70e501866becaded6b82075376fb48a18ec96a85", "messageHeadline": "Revert \"[TensorRT EP] reduce CI pipelines test execution time (#11440\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 0, "deletions": 10}}, {"node": {"oid": "69b2fab8109a8c8531d1d1a1f26029ddaa8c1abf", "messageHeadline": "Update DirectML from 1.8.0 to 1.8.2 (#11459)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "8467af832fea0e22daaac268f2a39921364ff7e3", "messageHeadline": "Fix reduced pipeline by excluding test case standalone op (#11458)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "3624f7c5a5d1bba95adca9e580e73daf808b4c45", "messageHeadline": "Update samples (#11420)", "author": {"name": "Brian Popow", "email": "38701097+brianpopow@users.noreply.github.com", "user": {"login": "brianpopow"}}, "additions": 25, "deletions": 20}}, {"node": {"oid": "2a90922f01e2fa9861dda0c7a769cfed09658167", "messageHeadline": "Using vectorized loads (float2) for fp16 to improve performance (#11390)", "author": {"name": "Hubert Lu", "email": "55214931+hubertlu-tw@users.noreply.github.com", "user": {"login": "hubertlu-tw"}}, "additions": 113, "deletions": 26}}, {"node": {"oid": "d2ae0f49b2cd7824613c27df7baa0dfa9559e1db", "messageHeadline": "Make Graph::InlineFunction be able to process initializers (#11443)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "8d6ade9e08fcac5c9ff0a96cddcc0db3d4da6abd", "messageHeadline": "[TensorRT EP] reduce CI pipelines test execution time (#11440)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "3f3c5fcd682cbc3f6682bbe3c79a68769c187002", "messageHeadline": "Unify the Compile API for mobile build and normal build (#10632)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 513, "deletions": 1158}}, {"node": {"oid": "eca4cbc419d0bc3c3ecdf4d2affa42f9bfe9fb56", "messageHeadline": "Avoid using word 'crazy' (#11396)", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "5ae461ec0a4c934b44b6939cc2332b9dbd4f2db2", "messageHeadline": "[TVM EP] update set input to remove excess copying inside TVM (#11247)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 16, "deletions": 8}}, {"node": {"oid": "084165c748bcd6fbb3fca4d60b1ece911c04c043", "messageHeadline": "Change MinGrad/MaxGrad to Use Distributed Logic (#11388)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 93, "deletions": 64}}, {"node": {"oid": "860ba8820b72d13a61f0d08b915cd433b738ffdc", "messageHeadline": "[js/rn] fix ORTRN for iOS (#11425)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 15, "deletions": 5}}, {"node": {"oid": "963e1ace4ed21fd7e368a97b077fc474ff551022", "messageHeadline": "Fix SAL annotations for custom op (#11432)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "a1f9847b239269e3f7262e3d2766799ecc2d38f7", "messageHeadline": "[Fix] Add the extra param to match gelu in PyTorch in the contrib sym\u2026", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 25, "deletions": 5}}, {"node": {"oid": "1aad59fa49c9cadbf134f4d0507a7ae9ac133113", "messageHeadline": "Increase timeout for IOS packaging pipeline (#11431)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "57b51e72d7d4d9b4a0d76c03dd1325940f551c5e", "messageHeadline": "Linux CI: uninstall onnx before installing it (#11428)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "af21a049772573de0100430dbef06488a2dc93b8", "messageHeadline": "[js] upgrade async@3.2.3 /js/ (#11421)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 102, "deletions": 87}}, {"node": {"oid": "85fa168dc1a9d4866472d3345f71a2fc2d99f979", "messageHeadline": "Add optional dft_length input to the DFT and IDFT operators. (#11427)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 188, "deletions": 125}}, {"node": {"oid": "ae043e3963eabf16522b0b1ef23531acc6a20977", "messageHeadline": "Support ort device tensor in ortmodule's inference (#11112)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 68, "deletions": 44}}, {"node": {"oid": "8d69b9398baf95e0c0781dd96e9fbb7f70294f4f", "messageHeadline": "APIs for custom op to invoke ort operator directly (#10713)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 949, "deletions": 29}}, {"node": {"oid": "a3e38d7c90462dabebe7ccb464e4f9f2b23982fe", "messageHeadline": "[js] upgrade async@3.2.3 /js/web/ (#11426)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "253c8b41edc54c35cfc798635db6d5daf56575e7", "messageHeadline": "Move some of the transpose kernel code to onnxruntime_framework.lib (\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 399, "deletions": 341}}, {"node": {"oid": "308b6050472450d390089d89b21a1926f6d3468f", "messageHeadline": "[wasm] increase timeout for Web Assembly static lib CI (#11306)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "d306e00351e353ac7f4498a04a52ad8a7d99c865", "messageHeadline": "[js/rn] set minSdkVersion to 21 for ORT-RN Android (#11403)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "5023f6750b1078383f60a17bd927ff306f2d4eac", "messageHeadline": "Revert \"Call pluggable EP's shutdown function in Environment::~Enviro\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 47, "deletions": 50}}, {"node": {"oid": "92cc1b4542d5fdc1ad00fcb1e14b7bf061a671c4", "messageHeadline": "Update GPT2 notebook (#11371)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 167, "deletions": 426}}, {"node": {"oid": "024747bff43147c75f5e9a53a6471248e3626a83", "messageHeadline": "Allow int32 as shape type (#11345)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "49d7050b88338dd57839159aa4ce8fb0c199b064", "messageHeadline": "Create Checkout Submodules Script (#11344)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 109, "deletions": 127}}, {"node": {"oid": "80a950cde196c9d150050e07a3a2e30d77e58e4b", "messageHeadline": "Create OV EP Signing Pipeline (#11391)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 33, "deletions": 0}}, {"node": {"oid": "4b875e354352ddc8e837800f429ae3c215fd6a7b", "messageHeadline": "Re-implment the function support in onnxruntime (#11167)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 1000, "deletions": 1077}}, {"node": {"oid": "e194a01787480878461a467ac7ed0136eac9e4cb", "messageHeadline": "Update SafeInt version. (#11379)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "11e681f46a0a1430702e0ddb708e5a5a6650c86c", "messageHeadline": "[ep-perf] Fix the order of onnx and onnxruntime imports. (#11383)", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c6a3c149de72492344c398fe8019e97ed6a42cbf", "messageHeadline": "Fix formatting. (#11386)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "c8270c2940f72a3722891fec7edf3c9ac542f904", "messageHeadline": "Add ATen export and gradient for torch.max/min (#11275)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 183, "deletions": 4}}, {"node": {"oid": "1c64351e091888e5737afec666f82b5c5c8873dc", "messageHeadline": "Create Tensor with Strides (#11294)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 67, "deletions": 49}}, {"node": {"oid": "6f218105ed495f3901943120c6122241db522cbd", "messageHeadline": "[FIX] amd ci pipeline timeout (#11387)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "833ded4b0e732ad3d620d076a5dfcb6617f1bb62", "messageHeadline": "Update setup.py to include config files used by model analysis in whe\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "6cd1931a935b3694f027184f52abad1e8064f571", "messageHeadline": "Specify list/map capacity when initializing where possible (#11110)", "author": {"name": "Mina Asham", "email": "mina.asham@hotmail.com", "user": {"login": "mina-asham"}}, "additions": 28, "deletions": 7}}, {"node": {"oid": "87a628a602d9e97aeb946380a08969f3055c9434", "messageHeadline": "[NNAPI EP] Add support for PRelu op (#11303)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 15, "deletions": 5}}, {"node": {"oid": "9e7753c109d4eb8e7efe5d98088dd9e6b7e12432", "messageHeadline": "Speed up GatherElements CPU Op (#11349)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 74, "deletions": 147}}, {"node": {"oid": "d64769c38e5ec60f1650e8b02ed7bdbd8c338291", "messageHeadline": "Set black's target version (#11370)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 26, "deletions": 25}}, {"node": {"oid": "f6526af23d622a33b0b64c6f1f8e1f999a8a1116", "messageHeadline": "Add LGTM config for c++ and c# (#11365)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 25, "deletions": 0}}, {"node": {"oid": "20ee399be47700a9bf44688147c3df391b54ebc0", "messageHeadline": "Include layout transformation ops in extended minimal build and above\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 170, "deletions": 59}}, {"node": {"oid": "88ff48910da946ebce8d695ea721e86ea9a13c7d", "messageHeadline": "Perf Bug Fix: Keep First Run for Random Input (#11366)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "a7d0158c24b7ba42eb782140826b0c3614f49981", "messageHeadline": "Introduce a way to disable Abseil library (#11353)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 177, "deletions": 50}}, {"node": {"oid": "63d4f45186653f1d28e52cb77d7f2463e7c5f438", "messageHeadline": "Add stub implementation of the NNAPI interface  (#11288)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 281, "deletions": 140}}, {"node": {"oid": "a937920e24f5653f887bb3a4219775b9de3fd004", "messageHeadline": "Update sequence_op.h (#11361)", "author": {"name": "Magnus1cheng", "email": "88537848+Magnus1cheng@users.noreply.github.com", "user": {"login": "chengchen666"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cbc7bd1d2a050ef0df7a4ec270b9fe6e6a2fc129", "messageHeadline": "Introduce parameterized as a dev dependency (#11364)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "14a2b2518e914a9e23c2fcb78f73d2104bdabec6", "messageHeadline": "[NNAPI QDQ] add qdq gemm/matmul op support (#11089)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 274, "deletions": 42}}, {"node": {"oid": "4ec5fe5c8af2b9e8d8bd512d7ac4d2a0809c4841", "messageHeadline": "Github action: Inline lint python / js / cpp (#11328)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 100, "deletions": 0}}, {"node": {"oid": "0292356bd7ce5b379be8ee9820bfbf4fd321ebc5", "messageHeadline": "TensorRT EP engine cache serialization/deserialization refactor (#11045)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 287, "deletions": 217}}, {"node": {"oid": "81d78706feb1dc923f3e43f7ba8ac30b55f5b19b", "messageHeadline": "Add location planning logic for implicit inputs which are graph input\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 180, "deletions": 14}}, {"node": {"oid": "fdce4fa6af437b0b822958ab47b3b8f77f9e14ae", "messageHeadline": "Format all python files under onnxruntime with black and isort (#11324)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 54748, "deletions": 36690}}, {"node": {"oid": "13f86e7d560fa3f94f5063a27fae396de44319fd", "messageHeadline": "print mac agent info (#11338)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 15, "deletions": 0}}, {"node": {"oid": "aaa583e776ef2dce8dc4748f1303a092e5b2e634", "messageHeadline": "Refactor the model tests code in onnxruntime_test_all.exe (#11300)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 257, "deletions": 143}}, {"node": {"oid": "6fb29f5b9a50f95e18c81a2052c00bac2f23e3dd", "messageHeadline": "Add python docstring linting in vscode settings (#11316)", "author": {"name": "Justin Chu", "email": "justinchuby@users.noreply.github.com", "user": {"login": "justinchuby"}}, "additions": 106, "deletions": 33}}, {"node": {"oid": "532e2536cc068ace0cfea9cd4d3f1308d9db0552", "messageHeadline": "increase timeout in PR build  (#11319)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "f069951835bf065917a516b827bad0b29939b368", "messageHeadline": "[trt-perf-test] Pass TensorRT/CUDA EP options via dictionary argument\u2026", "author": {"name": "Adrian Lizarraga", "email": "adlizarraga@microsoft.com", "user": {"login": "adrianlizarraga"}}, "additions": 124, "deletions": 47}}, {"node": {"oid": "ba1e9a218e68f3a6ec4e8c9303acb9fddcbdeea3", "messageHeadline": "increase timeout (#11310)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "daf87fd0dd3eaebb3ac9917797342770b9d6fbb0", "messageHeadline": "specify the path for gpt2_helper in onnx_exporter.py (#11301)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "23b01258b55b4438da36433a509f2b320628b7d2", "messageHeadline": "Fix how the output tensor is created in CUDA SpaceDepth ops (#11302)", "author": {"name": "Hariharan Seshadri", "email": "hasesh@microsoft.com", "user": {"login": "hariharans29"}}, "additions": 27, "deletions": 19}}, {"node": {"oid": "86cdabbcfd8e9cbbfd5bfe9da4541a7672390db3", "messageHeadline": "Add OpenVINO Pipeline Status to README (#11299)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4d0214f8510c740682082a00b00a5f2422142165", "messageHeadline": "Move Contains() helper function to a higher common.h. (#11289)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "7aa4af238ad46b815c83ff996a5567c865550ae5", "messageHeadline": "Add strict_shape_type_inference config option (#11081)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 287, "deletions": 186}}, {"node": {"oid": "c5de493a8a6a62d07f7ec32cebe784765826ede4", "messageHeadline": "Exclude EPs that aren't available on mobile to try and fix Xamarin bu\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "4854a09340260c3e75844268d594ece0bd0dd29f", "messageHeadline": "Consolidate utils::ToTensorProtoElementType, TypeToDataType, and data\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 109, "deletions": 227}}, {"node": {"oid": "2cacd18d51b243391d043ddf4e48e34d5f99fd30", "messageHeadline": "Fix an SAL annotation error", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "1d96cbec73a2febff99661f8187a07954555ac52", "messageHeadline": "Move gpt2 script to models\\gpt2 sub-directory (#11256)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 292, "deletions": 255}}, {"node": {"oid": "cb46d79108c953d25db29432245fa74544d4b44c", "messageHeadline": "Model tests refactor (#11194)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 17, "deletions": 9}}, {"node": {"oid": "af249943a184159997702a2fc284795ebc86cf53", "messageHeadline": "Increase the timeout so the packaging pipeline stops failing.", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "013306c940e12917348e405df959f30d1e54291d", "messageHeadline": "[MinBuild] 132KB minimal build binary size reduction via dummy __cxa_\u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 154, "deletions": 1}}, {"node": {"oid": "180b3f7cc2e2789a3a1093c6167c2430071207b5", "messageHeadline": "Update QDQFinalCleanup transformer to also handle removing DQ/Q node \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 252, "deletions": 128}}, {"node": {"oid": "e3ff4a6bfaf1f8a739317fae006f736bd487bacc", "messageHeadline": "Fix NNAPI EP error when handling external node adjacent to partition.\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 56, "deletions": 24}}, {"node": {"oid": "70d97bdf532502cb9da8ff8711fe3b0ff11cfec3", "messageHeadline": "Support only one input in QLinearConcat (#11265)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 31, "deletions": 5}}, {"node": {"oid": "2e6c2177af9c7a51c0d76d116e7569f415b69a85", "messageHeadline": "remove deprecated quantize api (#11263)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 80}}, {"node": {"oid": "acb555c4c7ba5a9314d6a4d562d3dfdb16f3816a", "messageHeadline": "ppc64le: Optimizing the MlasMaximumPool() to use VSX instructions (#1\u2026", "author": {"name": "Maxiwell", "email": "maxiwell@linux.ibm.com", "user": {"login": "maxiwell"}}, "additions": 36, "deletions": 0}}, {"node": {"oid": "bab9b80f1f2330d3a115e0abbb4d8278c2be3f44", "messageHeadline": "auto mixed precision for t5 (#11252)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 128, "deletions": 17}}, {"node": {"oid": "5ee8e2e491a5d3fba1100c1f652872b839577552", "messageHeadline": "[js] use NPM and yarn to upgrade package version (#11059)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 27, "deletions": 18}}, {"node": {"oid": "06026fe8e61160107a27bb727b3a747ded39d3a3", "messageHeadline": "SizeInBytes Fix for Strided Tensor (#11224)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 40, "deletions": 3}}, {"node": {"oid": "3dac66698b6a44ff860064907af33b60621743ca", "messageHeadline": "Add option to specify onnxruntime repo URL in tools/android_custom_bu\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "efb0928e2bba9d6e942abaad63a888a75dfc976f", "messageHeadline": "Fix find_package for benchmark", "author": {"name": "Lukas Berbuer", "email": "lukas.berbuer@gmail.com", "user": {"login": "lukasberbuer"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "98faaa7e2fe74191c9b038f594146165286717cb", "messageHeadline": "Scoped GIL release in run_with_iobinding (#11248)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "dec99657a1fefaf4046dc7f1832951ac379bd201", "messageHeadline": "Improve onnx shape inference in quant tool (#11106)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 99, "deletions": 74}}, {"node": {"oid": "9765ef8b4ea36fae40021b2c8451d9a63fa02e6e", "messageHeadline": "fix build warnings (#11213)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 37, "deletions": 33}}, {"node": {"oid": "0bad5b1b5ab9277108f3bbcbe5c62d60a67bbd5f", "messageHeadline": "[CUDA] Rollback TileMemcpy and TileBatchedMemcpy when Block Size is S\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 271, "deletions": 174}}, {"node": {"oid": "8ee8fdd59b7830bb0edde7c14067a73e57e2735e", "messageHeadline": "Add training api test runner (#10972)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 562, "deletions": 2}}, {"node": {"oid": "d9eeb483932934dca46071478bb57c61bdabe58f", "messageHeadline": "One dnn v2.6 update (#11220)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 126, "deletions": 79}}, {"node": {"oid": "227bc7264e16fee09ac349ade758e1552f09f3f7", "messageHeadline": "Fixed compilation error for ARM architecture (#11223)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "bc296c706ebd69e02c02f77f319710c984633266", "messageHeadline": "MatMulScaleFusion - handling scale input (#11121)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 55, "deletions": 1}}, {"node": {"oid": "94032357e2f137e34b6f2bb9c2c7a9a898faa513", "messageHeadline": "use int storage (#11185)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 46, "deletions": 9}}, {"node": {"oid": "63ff391b1642d9b479103f83c22753b01acce7a1", "messageHeadline": "add AppendExecutionProvider_CUDA_V2 to the C++ api (#11153)", "author": {"name": "Ahmad Zakaria", "email": "zakaria.mohammad@tomra.com", "user": null}, "additions": 6, "deletions": 0}}, {"node": {"oid": "c2b4054c7433dfe49d7f46bb744c8621bc411e38", "messageHeadline": "Fix typos", "author": {"name": "chausner", "email": "15180557+chausner@users.noreply.github.com", "user": {"login": "chausner"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "5216a43c9d0d07abf38ab9bd41c1bc81013afc09", "messageHeadline": "Consolidate TensorRT subgraphs to reduce inference overhead (#11211)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 24, "deletions": 5}}, {"node": {"oid": "a00d24066a3f7bd99ae2bec2dc6c1ebff522aa67", "messageHeadline": "Fix CITATION.cff and add automatic validation of your citation metada\u2026", "author": {"name": "Faruk D", "email": "144492+fdiblen@users.noreply.github.com", "user": {"login": "fdiblen"}}, "additions": 20, "deletions": 2}}, {"node": {"oid": "9707181257391323dfe7170d506650ded04dc59c", "messageHeadline": "fix build error (#11199)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "3b3b23bcf99486c1cfbe2f6afc495aaccb37a096", "messageHeadline": "Add new python helper dirs to wheel. (#11196)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "0d0edc071f24112dea9719096bd609b99954bfc0", "messageHeadline": "Detecting ARM64 CPU core micro-architectures in Windows (#11145)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 863, "deletions": 184}}, {"node": {"oid": "ddb17294b2424774b1c54fb3e3dba8f3496a8e9c", "messageHeadline": "Fix gradient builder for Cast (#11008)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 73, "deletions": 6}}, {"node": {"oid": "e84c338989244689656b19193de6a4ef2e6c38df", "messageHeadline": "minor improvements to CONTRIBUTING doc (#11080)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 14, "deletions": 20}}, {"node": {"oid": "5337972f92ac07327dbfee3bf99483dff9a27143", "messageHeadline": "Update to use teams instead of individual GH handles (#11163)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "9435369550532989f6bcc009b1e928cca5fdf587", "messageHeadline": "Option to build with DML as an external project (#11180)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 66, "deletions": 7}}, {"node": {"oid": "38e67e66a2ba8906bbdd644649441d893abeaabd", "messageHeadline": "Add script and Dockerfile to build custom Android package (#11144)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 206, "deletions": 6}}, {"node": {"oid": "e397d8e63e4e207ad102cdb999a3e3a1298f75d4", "messageHeadline": "POWER: Optimize MlasTranspose functions (#11172)", "author": {"name": "RajalakshmiSR", "email": "58744256+RajalakshmiSR@users.noreply.github.com", "user": {"login": "RajalakshmiSR"}}, "additions": 199, "deletions": 2}}, {"node": {"oid": "833f5d560494544c4b16568699096e642b15ae9b", "messageHeadline": "Remove dependancy on EP TVM in unit test project (#11170)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 0, "deletions": 9}}, {"node": {"oid": "625cc0ab990e878bb0f0724e51814c33753fe2a4", "messageHeadline": "Add Initialize() to shared providers to allow for reload (#11066)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 124, "deletions": 171}}, {"node": {"oid": "8237568b6585c38b6aceb7eaaded4cc1b9087ed8", "messageHeadline": "Fix the rocm packaging pipeline package upload problem (#11174)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 379, "deletions": 314}}, {"node": {"oid": "04fe1bd2ed8de3a11f148eb489fc24345b2ef21e", "messageHeadline": "Bump electron from 12.2.3 to 13.6.6 in /js/web (#10978)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "ae243c2bb5af11bc079b086eeb1a775a67cea92c", "messageHeadline": "Pull Nightly Wheel File and Cleanup Perf (#11164)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 71, "deletions": 165}}, {"node": {"oid": "749c0ddd1ea5b16a6c98414aeb8da61d5ac3f58f", "messageHeadline": "Upsample support NHWC (#10824)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 1116, "deletions": 349}}, {"node": {"oid": "269be2fe632c5ebf8b37b038fc7b36b4d4e6d95f", "messageHeadline": "Remove unnecessary option from convert_onnx_models_to_ort.py, fix old\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 15, "deletions": 19}}, {"node": {"oid": "00b595e389ebdb30c3ff2c2a261040b2e5907d08", "messageHeadline": "move longformer and t5 to models subdirectory (#11161)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 50, "deletions": 33}}, {"node": {"oid": "f24523e0eb11a918dd679f812feb85cce67fc77a", "messageHeadline": "Enable LayerNorm and SkipLayerNorm in OneDNN EP (#11128)", "author": {"name": "Erick Mu\u00f1oz", "email": "erick.munoz.alvarado@intel.com", "user": {"login": "eralmual"}}, "additions": 484, "deletions": 0}}, {"node": {"oid": "d96230065eb7eec492ecdc6492161ac43566754e", "messageHeadline": "fix code error in function.cc (#11148)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "12c687f594fd2a7c78c52a44ca70caa683fec55b", "messageHeadline": "Rework initializer.cc to eliminate code duplication (#11131)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 469, "deletions": 840}}, {"node": {"oid": "bcc62e0cbf6b9d5da9b38bbb0c88a485f4c8641a", "messageHeadline": "move some process out of training step (#11150)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 21, "deletions": 27}}, {"node": {"oid": "4c37f15c1ba3e24e0c7e2110f34ab410fadb2340", "messageHeadline": "Find boost, nsync, json, cpuinfo system libs with CMake option onnxru\u2026", "author": {"name": "Lukas", "email": "36054362+lukasberbuer@users.noreply.github.com", "user": {"login": "lukasberbuer"}}, "additions": 71, "deletions": 32}}, {"node": {"oid": "0aaf3a676aadf3f7dcf6545cd8fb04dc90b20a40", "messageHeadline": "Update reduce norm1/norm2 and layernorm kernels with ROCm 4.3.1 (#9399)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 2, "deletions": 133}}, {"node": {"oid": "1b664e6d4c83fd207425787fbfb90427e7395d9f", "messageHeadline": "Link cpuinfo only if supported (#11147)", "author": {"name": "Lukas", "email": "36054362+lukasberbuer@users.noreply.github.com", "user": {"login": "lukasberbuer"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "541eff8d893020ac9ba9a60b0833649ac829ad3a", "messageHeadline": "Directly use memory mapped data for external data initializers  (#11127)", "author": {"name": "IkerAriz", "email": "60264860+IkerAriz@users.noreply.github.com", "user": {"login": "IkerAriz"}}, "additions": 96, "deletions": 3}}, {"node": {"oid": "5637f171896158589f0e985291d5860c35daf37d", "messageHeadline": "Remove python frontend codeowners (#11143)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 0, "deletions": 12}}, {"node": {"oid": "760969446408c68e53ae753a32a092ffaf1aea86", "messageHeadline": "Enable building with a GDK (#11126)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 325, "deletions": 39}}, {"node": {"oid": "4983d6e5d697ed898380a6cf0cba8610d83de13f", "messageHeadline": "Call pluggable EP's shutdown function in Environment::~Environment() \u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 50, "deletions": 47}}, {"node": {"oid": "2700261f7ce7ea83749e73fcf414994d7e46d4c3", "messageHeadline": "Provide an API to supply external initializers data from user buffers\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 422, "deletions": 32}}, {"node": {"oid": "eec518780172038893eaccd8b745bfaa6573ca65", "messageHeadline": "Remove Rocm 4.2 from CI Build (#11130)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 5, "deletions": 258}}, {"node": {"oid": "e612018127358fa964da32d78cb082440771fe0d", "messageHeadline": "[CUDA] Tile Kernel Optimization (#11053)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 204, "deletions": 226}}, {"node": {"oid": "26fceca90f35fe94f39864ebaec0f2d51e367f10", "messageHeadline": "Update tools/ci_build/upload_python_package_to_azure_storage.py to no\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 27, "deletions": 53}}, {"node": {"oid": "81fa28bc56cc5630abfe72bf23e1024570d4058a", "messageHeadline": "OpenVINO-EP v4.0 Release PR with OpenVINO 2022.1 (#11025)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 1550, "deletions": 717}}, {"node": {"oid": "3f42665a40f1e3a655a9703ae668d903fbcd87d7", "messageHeadline": "Improve transfered time from ort to torch (#9610)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 673, "deletions": 76}}, {"node": {"oid": "58d97691ac6ef498c34989b0d109028bab596a82", "messageHeadline": "Set dims for constant with multiple values (#11116)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 62, "deletions": 8}}, {"node": {"oid": "91c940b619f1096f730ee1096c5befd4cc3e81d5", "messageHeadline": "adding fill scalar for torch ones direct initialization on ort device\u2026", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 51, "deletions": 3}}, {"node": {"oid": "2c2408814fe7db6e13c248617292597f1a58b8da", "messageHeadline": "Add function body for SoftmaxCrossEntropyLossGrad (#10779)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 486, "deletions": 52}}, {"node": {"oid": "20fbf603d300d2155dc0378da5c640f0ba40a978", "messageHeadline": "Fix ARM64EC build breaks (#11111)", "author": {"name": "Ben Niu", "email": "niuben003@gmail.com", "user": {"login": "mcfi"}}, "additions": 36, "deletions": 0}}, {"node": {"oid": "25fdf8b167dddb466021f09d7c999975b0a9cc4e", "messageHeadline": "Add Dequantize Linear operator on OneDNN EP (#11036)", "author": {"name": "Erick Mu\u00f1oz", "email": "erick.munoz.alvarado@intel.com", "user": {"login": "eralmual"}}, "additions": 334, "deletions": 0}}, {"node": {"oid": "8db180c245512f5bb5bdefc0f8b643ad6ff51294", "messageHeadline": "orttraining cuda 10.2 to not build for compute_80 (#11103)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "01631893cdeb92afdf7c16ffe2a9a7e64774dc48", "messageHeadline": "[cmake] Re-factor pre-compile header usage (#11093)", "author": {"name": "Jack\u00b7Boos\u00b7Yu", "email": "47264268+JackBoosY@users.noreply.github.com", "user": {"login": "JackBoosY"}}, "additions": 10, "deletions": 15}}, {"node": {"oid": "fc7fe0012fef5bd9c61900241071c297d98b2d42", "messageHeadline": "Fix: nodejs installer file name is wrong (#11097)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "872ed91d8a89ebb929afb9ea693721a06d6d2d28", "messageHeadline": "Perf FasterRCNN + MaskRCNN (#11102)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 13, "deletions": 1}}, {"node": {"oid": "112dec65655efb31c720d8b61d6254c3291caae0", "messageHeadline": "Added code for FusedMatMul inside matmul op primitive (#11077)", "author": {"name": "chethanpk", "email": "63478277+chethanpk@users.noreply.github.com", "user": {"login": "chethanpk"}}, "additions": 234, "deletions": 12}}, {"node": {"oid": "ea004e953feb7fe58ceb2ace705558b03f5414a2", "messageHeadline": "[cmake] Export multi targets in static build (#11063)", "author": {"name": "Jack\u00b7Boos\u00b7Yu", "email": "47264268+JackBoosY@users.noreply.github.com", "user": {"login": "JackBoosY"}}, "additions": 220, "deletions": 22}}, {"node": {"oid": "2dfd81b9bb097c90388010e5b7d298498274f8d9", "messageHeadline": "[cmake] Add option onnxruntime_ENABLE_CPUINFO (#11084)", "author": {"name": "Jack\u00b7Boos\u00b7Yu", "email": "47264268+JackBoosY@users.noreply.github.com", "user": {"login": "JackBoosY"}}, "additions": 29, "deletions": 23}}, {"node": {"oid": "25398cc5fe1d59fa280391c6a7887fef57013d28", "messageHeadline": "Add cleanup instruction to run_dockerbuild.sh (#11079)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 12}}, {"node": {"oid": "f9940f17b1e55833e0e42936f780539f211725c0", "messageHeadline": "Remove extra-index-url to avoid nuget security analysis vulnerability\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 12, "deletions": 9}}, {"node": {"oid": "b9279f637d12a66aae4cf285ba76e144a29106c3", "messageHeadline": "update How_To_Update_ONNX_Dev_Notes with right paths (#11074)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "588a66e2217083f61193b41d542dc5ea24c55599", "messageHeadline": "Add cleanup steps to the build jobs which run in Linux CPU machine po\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 63, "deletions": 74}}, {"node": {"oid": "249c4dec7f6c856e15cbfdfc7eb9ad89afbf6ec2", "messageHeadline": "Update orttraining release pipelines to use torch 1.11.0 (#11018)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 85, "deletions": 401}}, {"node": {"oid": "8e6dbad287f47b4d654b9a211deaa1e7faaa6fdf", "messageHeadline": "FIX: Nuget pipeline doesn't report binary size for Linux ARM64", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "11a4ca741da42231a892f158c6c6072994773c19", "messageHeadline": "fuse Conv+Add+activation for CPU from different op-branch (#10987)", "author": {"name": "wejoncy", "email": "247153481@qq.com", "user": {"login": "wejoncy"}}, "additions": 457, "deletions": 32}}, {"node": {"oid": "79e4ed8064cc002945f4d9bbd9dea8c234bfaf32", "messageHeadline": "Bump pytorch-lightning", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "eab7c0d5bfbe69895bfca6715d0aee216d0ed0a9", "messageHeadline": "Fixing optimizer failure due to missing provider list (#10497)", "author": {"name": "Boris Fomitchev", "email": "borisfom@users.noreply.github.com", "user": {"login": "borisfom"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "bfcd5bd4a2b77a8923cfc90c811a35e7607f4f43", "messageHeadline": "remove hardcoded library name (#11058)", "author": {"name": "Linnea May", "email": "linneamay98@gmail.com", "user": {"login": "linnealovespie"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8dcadba6701a2607c6b59f5eca426f81b3c7cfd1", "messageHeadline": "[js] aggregation of recent dependabot security warnings fix (#11060)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 121, "deletions": 167}}, {"node": {"oid": "e9c68d57ca9978a7016fbec2e94931d13a6d2238", "messageHeadline": "Bump minimist from 1.2.5 to 1.2.6 in /js/web (#11033)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "6c7090a82931be75b859dc44c506a2bd402a3608", "messageHeadline": "[js/web] fix output type mapping (#11049)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9505e8c6c134a597f7a3524ab84d724c2ca7ef66", "messageHeadline": "fix json format (#11046)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 10, "deletions": 5}}, {"node": {"oid": "9616ad483f5c67b36b4610a161e83e67373fb01c", "messageHeadline": "[Java] Support configuring CUDA and TensorRT execution providers (#10\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 682, "deletions": 9}}, {"node": {"oid": "179406bd253da7873c6e56bc7bfa18f3e9d1fa37", "messageHeadline": "[JS] upgrade package-lock.json from v1 to v2 (#11039)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 16390, "deletions": 139}}, {"node": {"oid": "998bf0fdb6b2e32fdb628f43c3c2b9bcdcd26195", "messageHeadline": "Remove advice to use IO Binding for this scenario (#11006)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "c37d2728bf6ba2c6638aea45cfcdc4da08004841", "messageHeadline": "Implement TreeEnsemble for opset(ai.onnx.ml)==3 (#10821)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1183, "deletions": 535}}, {"node": {"oid": "1424b796ff0af60b1d8b6e96fdc61515b41b305f", "messageHeadline": "[js/web] disable test_tan temorarily (#11048)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d1bdd2cd94e871cc29a63ac5cd37079fa9b4b742", "messageHeadline": "allow trailing slash in directory (#11001)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "5868413caf802bbd7f1dc0762b402988258d61d9", "messageHeadline": "fix seg fault (#11038)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 13, "deletions": 1}}, {"node": {"oid": "8f456735d1a3f5d45389ac2cbb9b00be86871275", "messageHeadline": "Remove unused variable. (#11043)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "6c005bfdbc28a04c521c364bc6b1df0f26e6acbc", "messageHeadline": "Enabled Cast operator on OneDNN EP (#11023)", "author": {"name": "Erick Alejandro Mu\u00f1oz Alvarado", "email": "erick.munoz.alvarado@intel.com", "user": {"login": "eralmual"}}, "additions": 236, "deletions": 0}}, {"node": {"oid": "6a6840d5c632360766d27849064f35fafbff842d", "messageHeadline": "Fuse LayerNormalization for Apex O2 (#10233)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 709, "deletions": 512}}, {"node": {"oid": "3b6cee80591fa001723b3d62b12b99149f11efe3", "messageHeadline": "[CUDA] Optimize Conv and ConvGrad for Training (#10999)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 70, "deletions": 24}}, {"node": {"oid": "8ba52b0a05da2351a0d8a4800449e919434f61f6", "messageHeadline": "Bump master version to 1.12 (#10797)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 27, "deletions": 21}}, {"node": {"oid": "9371401746b9b004a5ce14d0c38301e7d54e259a", "messageHeadline": "Move node EP assignment for ORT format into SessionState::FinalizeSes\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 194, "deletions": 214}}, {"node": {"oid": "9c6cc018a9a71f2d3b36647b83ef60659ebb2a4c", "messageHeadline": "Add utility to get the gradient graph from GradientGraphBuilder (#10995)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9, "deletions": 4}}, {"node": {"oid": "dc72159105ec6dadcb33699e04cdc25d1158d422", "messageHeadline": "Symmetric Quant indirect Conv kernel for ARMv8 A55 chip (#10862)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1569, "deletions": 330}}, {"node": {"oid": "8ddc45f52d6ad92d3d9aae1bb5cb01d7f03d3244", "messageHeadline": "Add linux and macos arm64 java aritifacts (#10981)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 22, "deletions": 0}}, {"node": {"oid": "d1be71eaa3892640eb30fc846fafb7ffe767d05d", "messageHeadline": "[cmake] Add keyword STATIC to add_library in function onnxruntime_add\u2026", "author": {"name": "Jack\u00b7Boos\u00b7Yu", "email": "47264268+JackBoosY@users.noreply.github.com", "user": {"login": "JackBoosY"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cb31b7eab191bc6a0f9e051f136a18b709126614", "messageHeadline": "Fixed creation of ORT_Value to pass offset of 0 (#11004)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "47c09e670126b33a57491627006d7d447754373e", "messageHeadline": "Clarify usage of kOnnxDomainAlias. (#10962)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 60, "deletions": 57}}, {"node": {"oid": "89ef987ab1cddd0efab0fd1f1e27738da7befe92", "messageHeadline": "Improve NonZero on CUDA/ROCM (#10307)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 222, "deletions": 163}}, {"node": {"oid": "1e917c879e3875d25413ac6586fa07f12262b709", "messageHeadline": "Adding support for saving and loading train step info properties in t\u2026", "author": {"name": "mpapdiwala", "email": "99221944+mpapdiwala@users.noreply.github.com", "user": {"login": "mpapdiwala"}}, "additions": 90, "deletions": 2}}, {"node": {"oid": "989e6400094ea39675fd6675ec096707800247b8", "messageHeadline": "Update docstrings in quantize.py (#10952)", "author": {"name": "Christoph Hausner", "email": "64791786+chausner-audeering@users.noreply.github.com", "user": {"login": "chausner-audeering"}}, "additions": 5, "deletions": 8}}, {"node": {"oid": "3c5853dcbc9d5dda2476afa8c6105802d2b8e53d", "messageHeadline": "register custom_op_symbolic for squeeze (#10970)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 41, "deletions": 0}}, {"node": {"oid": "7ee52fb8a0524b688be2bfec25b94512b2bde597", "messageHeadline": "amdmigraphx_ep-add ops to be supported by migraphx and fixed a bug in\u2026", "author": {"name": "Shucai Xiao", "email": "shucai@gmail.com", "user": {"login": "scxiao"}}, "additions": 440, "deletions": 190}}, {"node": {"oid": "ae08f9666dfd779a2ef0cf5b6071f0d89575af3e", "messageHeadline": "Fix type constraints in registration of DequantizeLinear (#10986)", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "938f3857a5a61020794a06a74c42ac4ca1028b4d", "messageHeadline": "Set the default for the STFT onesided attribute to 1, which tests exp\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "07201726ed364d8d6abda84d37ad689b5a9289b5", "messageHeadline": "Fixed macros for graph transformer registration. (#10983)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "de384805cd1ca2e94fc88942165e8a8d0ceb54b7", "messageHeadline": "Custom parameters (#10964)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 71, "deletions": 36}}, {"node": {"oid": "9a3be9b46adb8ada41a759229935ba30cb233267", "messageHeadline": "use #include <hiprand/hiprand.h>, not deprecated #include <hiprand.h>\u2026", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0efbe92296f7793d4e05a9d27bf9f25d05509e22", "messageHeadline": "fix coverage report error in master build (#10969)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 93, "deletions": 20}}, {"node": {"oid": "480c793125c67e3c873f8083287571bf04f42f52", "messageHeadline": "Update training packages to Pytorch 1.11.0 (#10851)", "author": {"name": "raviskolli", "email": "48601275+raviskolli@users.noreply.github.com", "user": {"login": "raviskolli"}}, "additions": 58, "deletions": 60}}, {"node": {"oid": "565318ce86472540d71c79223c5fe474d7eddb02", "messageHeadline": "Support ORT WASM compilation with the training flag (#10973)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "b28e5064f374f4298559ff64b7f534d42b42e75b", "messageHeadline": "Ignore DequantizeLinear nodes in CommonSubexpressionElimination optim\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 31, "deletions": 1}}, {"node": {"oid": "b88fb68fac3e6ce132c9cd2f9689ccb1afdf973e", "messageHeadline": "Adds missing numpy type when looking for the ort correspondance (#10943)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 64, "deletions": 35}}, {"node": {"oid": "dce5d719c5007bc56b5ecd2ba0fe8daf44771f16", "messageHeadline": "add build flag for emscripten settings (#10963)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 43, "deletions": 10}}, {"node": {"oid": "027565b3b2e2d84d9d61d4855117715d6cb818a1", "messageHeadline": "Add multi-dim dft test, and fix complex idft (#10947)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 252, "deletions": 60}}, {"node": {"oid": "2da82fd0b9a80d012521760091d5d7a503d86730", "messageHeadline": "allows multiple '--cmake_extra_defines' flags (#10953)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 21, "deletions": 21}}, {"node": {"oid": "6d19c295d0ae499d71b678ede309659397a2310c", "messageHeadline": "use lf as eol for node package (#10965)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "b34d9f6867850d309b4e8c407c8f3b39c514656f", "messageHeadline": "[js/wasm] Add WebAssembly static library build into web CI pipeline (\u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 66, "deletions": 41}}, {"node": {"oid": "4a5b5328a46d1a5a869a7cbeadd821fdddc70137", "messageHeadline": "Added support to Eager CodeGen for multiple in-place parameters. (#10\u2026", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 81, "deletions": 40}}, {"node": {"oid": "1cc2cfb7b807632a5ec5f5ff8fa8cfe34584326c", "messageHeadline": "Move #ifndef ORT_CXX_API_THROW to the no exceptions case. (#10937)", "author": {"name": "Leandro Gracia Gil", "email": "leandro.gracia.gil@gmail.com", "user": {"login": "leandro-gracia-gil"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "a6ea278502e757049312eb6b11865a6bd2e7d07b", "messageHeadline": "add python3.10 support  (#10848)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 51, "deletions": 11}}, {"node": {"oid": "8703d375170984a21cc19bab9a9acf8074f0a629", "messageHeadline": "Extend DropoutGrad function to support bfloat16 (#10662)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 38, "deletions": 28}}, {"node": {"oid": "91722e2bc4ef63add6d548f2b1212a7fd0b10531", "messageHeadline": "Fix typos (#10935)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c1e37e4ebf04008a41fa4f459c2f936a1bbd2d52", "messageHeadline": "Android CI Pipeline: Fix post coverage bug (#10949)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "fe6ab719f3bb389c7368c6be28c1b38884f603cb", "messageHeadline": "Fix a typo in quantization tools (#10940)", "author": {"name": "Ella Charlaix", "email": "80481427+echarlaix@users.noreply.github.com", "user": {"login": "echarlaix"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "eabb14788a7f1388d1b1f2c0da6cab0b2c480503", "messageHeadline": "[perf_metric] added inferences per second metric (#10921)", "author": {"name": "soundarthiaga", "email": "90360019+soundarthiaga@users.noreply.github.com", "user": {"login": "soundarthiaga"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "3897b93606c498c085cd69ca9a5d4c269394890b", "messageHeadline": "optimize Android CI (#10938)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 254, "deletions": 29}}, {"node": {"oid": "2dea7dc27f8d4382a042a0e6487fc925d36223fc", "messageHeadline": "Skip python arena shrinkage test on ppc (#10901)", "author": {"name": "Kotaro Yamamoto", "email": "kota.crk@gmail.com", "user": {"login": "kota-row"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "de06d950964d76285d6a3b4048604e63b68333b8", "messageHeadline": "[parallel_inference] added support for parallel inference with timed \u2026", "author": {"name": "soundarthiaga", "email": "90360019+soundarthiaga@users.noreply.github.com", "user": {"login": "soundarthiaga"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "5cbacec854f151549f13b4be0a19c65c01a5728d", "messageHeadline": "Maintain aspect ratio by doing resize + crop in image_to_pb tool (#10\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 23, "deletions": 4}}, {"node": {"oid": "f058c59407b891967e5372e5e8b0862322e73fb7", "messageHeadline": "Performance: add io_binding support for bert benchmark util (#10907)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 63, "deletions": 5}}, {"node": {"oid": "c29d94383d7a7d66ac8ad92076d4f0b6a6f58dea", "messageHeadline": "Update CODEOWNERS (#10932)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 13}}, {"node": {"oid": "ee05c591e528caa46ceac37952d3f7645e859b33", "messageHeadline": "Fix benchmark bugs and add Pytorch version control  (#10928)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 105, "deletions": 24}}, {"node": {"oid": "6f844522c8c5c2b1125dbfe9de99e1ddfb21a470", "messageHeadline": "Follow up update for python API checking if `vcruntime140_1.dll` is a\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "8860fded02435e62b63bf69f3d420998f6a00f39", "messageHeadline": "Disable Some Einsum ORTModule Tests Due to Issue from PyTorch Exporte\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 27, "deletions": 6}}, {"node": {"oid": "5ed2f4ad5ffda8d4f71bf519a08b03c96bc40b16", "messageHeadline": "Remove Windows Store specific code", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 21, "deletions": 80}}, {"node": {"oid": "a3eeb7b0f7e53b648af4058486edc744733fd03d", "messageHeadline": "Update orttraining-linux-external-custom-ops.yml for Azure Pipelines", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "78133434b58337aef536aeb06b248deccf9f564a", "messageHeadline": "Fix fp16 converter bugs[1/n] (#10882)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "5d4ff67c362895a324d502eed14af255502e9f80", "messageHeadline": "Support fusion options for benchmark.py (#10900)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 24, "deletions": 12}}, {"node": {"oid": "b86d105153ec51c2db908595bf80f48041d9979b", "messageHeadline": "[python API] Change raise import error when `C:\\Windows\\System32\\vcru\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "740870a285229fa3e2a8c831c8ce56273e1ff67d", "messageHeadline": "[js] Create npm packaging pipeline (#10886)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 418, "deletions": 283}}, {"node": {"oid": "766e6ac4fdfa46e5b7e2460d08634f88ce14dc4f", "messageHeadline": "check TVM target for CPU (#10926)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "a46b00499a80c5c3b67cfe941bc54178cf8e011b", "messageHeadline": "Revert \"Upsample support NHWC (#10554)\" (#10917)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 275, "deletions": 342}}, {"node": {"oid": "810c18e809f03baddab58d3c9161295d36e6d7c2", "messageHeadline": "fix complex multi-dim dft (#10896)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 60, "deletions": 15}}, {"node": {"oid": "07a71d5bf231fce2995be2075bdda9b3ed9bcfb9", "messageHeadline": "Fix handling of nodes inserted by NHWC transformer. (#10904)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 9}}, {"node": {"oid": "e03b799b957a1828b2c470213ca3feac7edee78a", "messageHeadline": "Make it clear in verbose log on why a kernel impl does not fit for a \u2026", "author": {"name": "cloudhan", "email": "guangyunhan@microsoft.com", "user": {"login": "cloudhan"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "6c0eff1ae4055cd1261582894a169e863dffe1d2", "messageHeadline": "optimize Transpose3DKernel (#10891)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 54, "deletions": 52}}, {"node": {"oid": "463fac67a383611a507a0138a17e1403bd2cc04b", "messageHeadline": "[FIX] symbolic shape infer error with onnx-1.11.0 (#10674)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 24, "deletions": 13}}, {"node": {"oid": "d7d7665023e29d686d462c9a304e5ca376904c71", "messageHeadline": "restore random states after export_model (#10705)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 65, "deletions": 0}}, {"node": {"oid": "42d7112f03e59ea7f1fb42d0fe4954f5f8ec950e", "messageHeadline": "Add multithreading test and put a lock on nvinfer1::createInferRuntim\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 200, "deletions": 14}}, {"node": {"oid": "ce204d07443951f986ce8695790da2418ac25c34", "messageHeadline": "Update to flatbuffers v2.0.0 (#10866)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 35, "deletions": 27}}, {"node": {"oid": "bac9c0eb50ed5f0361f00707dd6434061ef6fcfe", "messageHeadline": "skip optional related models from opset16 (#10840) (#10878)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "576365771547c2fecfad272efc93b090b545b6bb", "messageHeadline": "[UPDATE] Add prefix in front of the file (#10884)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 30, "deletions": 0}}, {"node": {"oid": "625a1f7673fb7a42975d85e850df30ea835c49ad", "messageHeadline": "[TVM EP] code refactor (#10655)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 979, "deletions": 645}}, {"node": {"oid": "f468ea40e59c6f3d7ca06c3d0c0b6d325c185602", "messageHeadline": "Refactor Node::AddAttribute() (#10869)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 255, "deletions": 265}}, {"node": {"oid": "040c0645e2ce600df1284d314976f797fc5f61c0", "messageHeadline": "[ADD] Add micro-benchmark for Cast (#10870)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 105, "deletions": 0}}, {"node": {"oid": "860f28254e8a14de7a490833e612d011678c8440", "messageHeadline": "Update DFT definition to more closely align with PyTorch by enabling \u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 368, "deletions": 161}}, {"node": {"oid": "de6d1fcb411e80525834e237551f9e3fbfab76e0", "messageHeadline": "Update C# runtest.sh", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "aebbb90b7945834da48c3ca7e5d7c6c990211088", "messageHeadline": "Integrate C-API tests into Pipelines for release packages (#10794)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 71, "deletions": 15}}, {"node": {"oid": "2d961604b16fe3966ea75aec897492202a6cae5a", "messageHeadline": "Refactor Python API docs to better explain IO binding scenarios (#10651)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 122, "deletions": 73}}, {"node": {"oid": "f385c73058176ed81493319d3e49b5cc9382f63f", "messageHeadline": "Fix a couple of issues with the python package tools (#10858)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 57, "deletions": 3}}, {"node": {"oid": "0d8d44d035c1587048c0bf22c0550588ae93a7c2", "messageHeadline": "Fix Reduced ops pipeline (#10861)", "author": {"name": "Hariharan Seshadri", "email": "hasesh@microsoft.com", "user": {"login": "hariharans29"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "7e9dfe627a30591001ecccae212eeb2a326c8246", "messageHeadline": "Add additional NNAPI QDQ test cases for expected failure path (#10769)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 116, "deletions": 32}}, {"node": {"oid": "e53422c6d0ff6f44cb96a3a706eebcd5d8850d11", "messageHeadline": "Update convert_onnx_models_to_ort.py to support runtime optimizations\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 403, "deletions": 255}}, {"node": {"oid": "03181caeaef9343bd2e8d7999386893562ebb23a", "messageHeadline": "Creating test case for printing ort tensor (#10850)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "ce10d7d23130981469d8851c822d61152f88cfac", "messageHeadline": "Workaround https://github.com/microsoft/STL/issues/434#issuecomment-9\u2026", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 15, "deletions": 8}}, {"node": {"oid": "d1f059c856a96746a5c526145c97b5522189ef21", "messageHeadline": "Remove kernel32 defaultlib", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 16, "deletions": 10}}, {"node": {"oid": "ddf1a312ab92d8eb7fb18ea09e2f75a0b452e63e", "messageHeadline": "Remove cpuinfo from WCOS builds", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c4f73af234efb8cbe3df3e4dd14c19313bc9b06a", "messageHeadline": "Fix wrong percentile values returned during calibration (#10847)", "author": {"name": "Funtowicz Morgan", "email": "mfuntowicz@users.noreply.github.com", "user": {"login": "mfuntowicz"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "6ac3b8d46a6efdfdf74c334bc8b8e9aadf639572", "messageHeadline": "Add support for opset 16 to transpose optimizer. (#10841)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 67, "deletions": 28}}, {"node": {"oid": "2d2eebb844a8a290bdcf67b7e9c4f9c1c6985539", "messageHeadline": "Correct a comment", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2e7592ddf82eebda97edfee69381f94e51014b05", "messageHeadline": "avoid using LocalFree on FormatMessageW buffer (#10796)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 8, "deletions": 10}}, {"node": {"oid": "64556888a10e3048b85ae2565d338aaded8e9e78", "messageHeadline": "add python binding for RunOptions config entry (#10694)", "author": {"name": "Kotaro Yamamoto", "email": "kota.crk@gmail.com", "user": {"login": "kota-row"}}, "additions": 48, "deletions": 1}}, {"node": {"oid": "d478a53d433473cad296bcdae2edc43a0fef210c", "messageHeadline": "don't clear grad_fns & add test (#10671)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 120, "deletions": 59}}, {"node": {"oid": "1a62306db70378d166dd255864de8d7cba18956b", "messageHeadline": "Use separate build directories for full and mobile iOS packages. (#10\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "5202efd11eab6c56c07d98c094b5f1b17b7ccfcd", "messageHeadline": "remove unused six in code and CIs (#10832)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "f87a06cd9617ab67d06e57e4c867e799f0d1c255", "messageHeadline": "Patch absl so that it doesn't disable important VC++ warnings (#10836)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 41, "deletions": 0}}, {"node": {"oid": "97ae44d06072b3e8fcf8cc7d3bd4b8b13d046a88", "messageHeadline": "Mark end of version 11 C API. (#10803)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3ae2bfaefe4344820d0a17055aa23f080530af34", "messageHeadline": "Abjindal/torch api change gelu (#10833)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "1d545dfe87594a78ea5b02ac4335d0fd22261315", "messageHeadline": "Address performance issue with abseil flat_hash_table. (#10819)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 28, "deletions": 28}}, {"node": {"oid": "e80ff632747b1eb8aaee2c7d600fb3a662041ee4", "messageHeadline": "Fix bug in MemcpyToHost (#10816)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 75, "deletions": 14}}, {"node": {"oid": "9853eaa14ff857ff28c153b2aaee87e461a3a919", "messageHeadline": "Detect runtime CUDA JIT and warn the user (#10781)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 21, "deletions": 0}}, {"node": {"oid": "cc3a3476ed866013faacfb2b53f8c3a98315437e", "messageHeadline": "Uninstall onnxruntime-training before running local tests (#10827)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 65, "deletions": 52}}, {"node": {"oid": "9cbcc93e03b85bc4c935408942e39c428b3e63f3", "messageHeadline": "Add micro-benchmarks for Attention and SkipLayerNormalization ops. (#\u2026", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 309, "deletions": 25}}, {"node": {"oid": "1c313f447656d759d7d8cad1dfa15b06c9cfd33e", "messageHeadline": "changing gelu backward op and adding required files (#10813)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "0293e525ead84d3cf3d8481ca27c9c480fc89d70", "messageHeadline": "Make QDQSelectorActionTransformer() is_int8_allowed parameter require\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 16, "deletions": 13}}, {"node": {"oid": "cc6bc34c8cdc8aa0e37764d2ae219a932024f1ca", "messageHeadline": "Update protobuf submodule (#10801)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "58521fb822aeea3e6ee3274be06b2f63a9ae4b88", "messageHeadline": "Make training CUDA kernels to adhere established code structure patte\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 422, "deletions": 272}}, {"node": {"oid": "4ef81b142dbd29abc3c64797084b8e61cf3212d1", "messageHeadline": "Making the Java tests faster by optionally disabling ones which requi\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "ae97ecf05b0463ae272025b213264425147958b9", "messageHeadline": "Fix CPU, CUDA Selu activation logic (#10771)", "author": {"name": "Hariharan Seshadri", "email": "hasesh@microsoft.com", "user": {"login": "hariharans29"}}, "additions": 11, "deletions": 2}}, {"node": {"oid": "c147c9dda635f281cdacaf8b6417eec04f7fede4", "messageHeadline": "Remove ORT_ENABLE_RUNTIME_OPTIMIZATION_IN_MINIMAL_BUILD. (#10778)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 371, "deletions": 399}}, {"node": {"oid": "769aa8363dd7ac0e6601e55481f7a9938ebc7c5a", "messageHeadline": "update onnx-tensorrt to bring in https://github.com/onnx/onnx-tensorr\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f4fd67cc2c55730ef15e589870eec8a3d19b16ca", "messageHeadline": "Revert \"add load from buffer (#10162)\" (#10590)", "author": {"name": "Jingqiao Fu", "email": "jingf@microsoft.com", "user": null}, "additions": 4, "deletions": 116}}, {"node": {"oid": "7e04dccca7a11d65c4ea1784d18cad5fc7867440", "messageHeadline": "Bump numpy in /tools/ci_build/github/linux/docker/scripts (#10385)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "68c8f5a1ef7dad5f5422d6d122201f342345433f", "messageHeadline": "Change a pipeline vmImage from windows-latest to windows-2019 (#10804)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "33c681919623ec850c3971f79cb53ded44d8b885", "messageHeadline": "add qdq support of Sigmoid (#10800)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 71, "deletions": 1}}, {"node": {"oid": "626073353343e66f39f4a6c1e4a56c8a1d5c4a9c", "messageHeadline": "Fix eager mode pipeline (#10802)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 7}}, {"node": {"oid": "a9d9c6b486720d55709c36ed07c14782e5fc96df", "messageHeadline": "Register CPU, CUDA and ROCM opset-16 kernels for some operators (#10643)", "author": {"name": "Hariharan Seshadri", "email": "hasesh@microsoft.com", "user": {"login": "hariharans29"}}, "additions": 463, "deletions": 178}}, {"node": {"oid": "ce07dc30fd5f90bde20b40ee6d905055a31630a7", "messageHeadline": "Change how we apply patches to absl (#10799)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "1e4a4bfe58de18d07a66a458ff664cb3d998cdcb", "messageHeadline": "update onnx-tensorrt reference. (#10795)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "da885a72e831a83bba47872470c29053d44bfd1f", "messageHeadline": "update with onnx 1.11 release (#10441)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 853, "deletions": 599}}, {"node": {"oid": "80917342b7c77d7661e230a27ea188230f6655fc", "messageHeadline": "[js] upgrade mocha@8.2.1 to 9.2.1 (#10793)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 316, "deletions": 218}}, {"node": {"oid": "4d943c9bd3e1a80af7ce3c163160082f62815a5d", "messageHeadline": "Bump numpy from 1.16.6 to 1.21.0 in /tools/ci_build/github/linux/dock\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "c07a27a00843edb669d6de0da3b789ac8975ff3e", "messageHeadline": "[FIX] delete python3.6 from AMD python package docker image builder (\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "4a38f9e31d916e5866203bf5ce0489b187e9f8fe", "messageHeadline": "enable strided tensor for training only (#10748)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 77, "deletions": 19}}, {"node": {"oid": "b7f00b9682256c0baf291de8cf620bb5656faad1", "messageHeadline": "Refactor the common code per operator into an abstract base class. (#\u2026", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 191, "deletions": 160}}, {"node": {"oid": "a08036da099ec512bdc6e115c86fce228ad9c87c", "messageHeadline": "correct symbolic name of GridSample operation (#10782)", "author": {"name": "Daigo HIROOKA", "email": "daigo.hirooka@gmail.com", "user": {"login": "daigo0927"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3e54f94bb0f6f8cc7f81493d28ace0afeaacb61f", "messageHeadline": "Bump karma from 6.3.14 to 6.3.16 in /js/web", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 30, "deletions": 29}}, {"node": {"oid": "25fdcfbd1499d123798e32d7b32e70c74b6ffc05", "messageHeadline": "[js/web] allow multiple inference session creating concurrently (#10784)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 26, "deletions": 8}}, {"node": {"oid": "a4b5fa334aa939fb159bdc571ed3d56ca8d31fc7", "messageHeadline": "Add type and shape information to profiled numbers (#10773)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 33, "deletions": 7}}, {"node": {"oid": "d8bf9a479bc41fa54c4d8bbeac33045851051844", "messageHeadline": "Remove python 3.6 from training pipelines (#10780)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 15}}, {"node": {"oid": "9d30262422321fe12d33a38e44f890438396f535", "messageHeadline": "Fix AMD training pipeline (#10788)", "author": {"name": "Hariharan Seshadri", "email": "hasesh@microsoft.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "50a6f095cd2a13e524fba054cf82cfc76fa57185", "messageHeadline": "Symmetric QGEMM kernel for ARMv8 A55 chip (#10754)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 955, "deletions": 3}}, {"node": {"oid": "55af7a96a729471720549d86382aefcc7a37ef2e", "messageHeadline": "update the amd ci pipeline (#10723)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 182, "deletions": 117}}, {"node": {"oid": "60acfd3dd8dabe082f96fa49fab7278a6fd33eb4", "messageHeadline": "Support CUDA Graph in the CUDA EP (#9978)", "author": {"name": "Fei Hu", "email": "hufei68@gmail.com", "user": {"login": "feihugis"}}, "additions": 653, "deletions": 72}}, {"node": {"oid": "0e335aba379708593e3f2080262d7c8bceea0665", "messageHeadline": "Update BeamSearch operator spec to support t5 (#10777)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 150, "deletions": 57}}, {"node": {"oid": "6be5185088839aace72c80133d13503c45fc5c8c", "messageHeadline": "Update dnnl Add, Mul, Sub, Div ops to handle scalar values (#10756)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 22, "deletions": 38}}, {"node": {"oid": "259ade25576fd53ff48f268f895f546648612a84", "messageHeadline": "Add ability to modify num_hidden_layers from benchmark script (#10760)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 89, "deletions": 40}}, {"node": {"oid": "fde847473b91c7feeaefc1ee641306a8147fcf0b", "messageHeadline": "Add min max moving average calibration method (#10753)", "author": {"name": "Ella Charlaix", "email": "80481427+echarlaix@users.noreply.github.com", "user": {"login": "echarlaix"}}, "additions": 43, "deletions": 8}}, {"node": {"oid": "43ff27c7c8949fa0b12323f0a5a284cb84b709b2", "messageHeadline": "ppc64le: optimizing the MlasQuantizeLinear() with VSX (#10644)", "author": {"name": "Maxiwell", "email": "maxiwell@linux.ibm.com", "user": {"login": "maxiwell"}}, "additions": 324, "deletions": 3}}, {"node": {"oid": "379b3cdef6630ef1a664a3ec797262979ef00977", "messageHeadline": "T5 to ONNX conversion script (#10766)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1087, "deletions": 0}}, {"node": {"oid": "12eb660415484505eb46b30d1929cb3ae5036613", "messageHeadline": "Compare TRT vs ORT-TRT Accurately (#10565)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 92, "deletions": 92}}, {"node": {"oid": "e3c85d4262ff9dc3675abf826d5f5cd7f76517aa", "messageHeadline": "Bump numpy", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "b780a3784e018194c624febb01d388334599b499", "messageHeadline": "Bump numpy in /tools/ci_build/github/linux/docker/scripts/training", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0b0e8ccf929ec08f9b5ede251d56704417eb4efe", "messageHeadline": "Bump numpy", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "283d0c47b44d7e6010cfc4e143b89efee9233653", "messageHeadline": "Update our absl cmake files (#10762)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 23, "deletions": 35}}, {"node": {"oid": "4c88fa59713a4fa8f65506b840b35dc3a859a83c", "messageHeadline": "Add micro-benchmark for FastGelu (#10744)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 136, "deletions": 13}}, {"node": {"oid": "46d0b20ac28aafae5edac843bb5b27a7b14eaeb0", "messageHeadline": "upstream TVM. small code cleaning (#10515)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "395a7242d6e360a7054d53b01c0e25d84ec4b7c0", "messageHeadline": "[iOS packaging] Minor updates. (#10755)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 34, "deletions": 28}}, {"node": {"oid": "e337f5faf3ec28e401113941bb0b118dcb18ae29", "messageHeadline": "Enable QDQ cleanup and NHWC optimizers in an extended minimal build. \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 422, "deletions": 253}}, {"node": {"oid": "7aa706854f6d5a4438aa40488cc5d988bb3bd884", "messageHeadline": "Pipeline changes to build full ORT package for Android (#10654)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 56, "deletions": 21}}, {"node": {"oid": "6072c6b65e9a8816ba53ccbdb15dbd7fc7001e18", "messageHeadline": "Simplify QLinearConv registration so type reduction works with it. (#\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 98, "deletions": 87}}, {"node": {"oid": "c2c85dd6b1daa04c96dff74581f4d5540a42377e", "messageHeadline": "Add an option to export ONNX graphs in ORTModule tests (#10579)", "author": {"name": "Abhishek Kulkarni", "email": "abd.kulkarni@gmail.com", "user": {"login": "adk9"}}, "additions": 18, "deletions": 6}}, {"node": {"oid": "745fa5885fbd7d3ef34dd153663ec5adb94110fc", "messageHeadline": "optimize web assembly build flags for multi-thread (#10759)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 10}}, {"node": {"oid": "c8ec7782bdc64237d6b8102b0bbce35d941a8a27", "messageHeadline": "Fix unused variable warning, move variable definitions closer to usag\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 9, "deletions": 10}}, {"node": {"oid": "ed87e1b7216d234cd270a9963417bd22297f9ebd", "messageHeadline": "Change axis to 0D in cumsum tests. (#10715)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 31, "deletions": 26}}, {"node": {"oid": "b3e96d61951a9ae05abec64b08643e1310fb4166", "messageHeadline": "A new pipeline to replace the existing WindowsAI packaging pipeline (\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 548, "deletions": 0}}, {"node": {"oid": "fe8d867efadf336bdd9dac9241d74a3ae6b45b6b", "messageHeadline": "Optimize BinaryElementWise and BiasGeluGrad kernels for AMD (#10594)", "author": {"name": "Hubert Lu", "email": "55214931+hubertlu-tw@users.noreply.github.com", "user": {"login": "hubertlu-tw"}}, "additions": 38, "deletions": 15}}, {"node": {"oid": "4c20f6863dfc1eeb50223cb6e9db68c3c8491d4f", "messageHeadline": "Fix build with gcc 7.5 (#10567)", "author": {"name": "cloudhan", "email": "cloudhan@outlook.com", "user": {"login": "cloudhan"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "75160d677975bf0063e3ef3b16c451e920781ed2", "messageHeadline": "Add the missing status return in beam search (#10738)", "author": {"name": "Fei Hu", "email": "hufei68@gmail.com", "user": {"login": "feihugis"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "a9dc50ba8b324e4f653eff2b172421d8b2e1b3c3", "messageHeadline": "Add option to force QDQIsInt8Allowed to return true when exporting to\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 31, "deletions": 9}}, {"node": {"oid": "44d08d80a03654b4e8e06d8a49056258a2050be8", "messageHeadline": "Add restriction to first usage in allocation planner (#10724)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 118, "deletions": 1}}, {"node": {"oid": "47ab0c200642815234942f01b84251add3ecaf6a", "messageHeadline": "Auto mixed precision conversion of GPT-2 onnx model (#10711)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 179, "deletions": 28}}, {"node": {"oid": "7ebff2b273366d0f44329369d95f50fd59ca9b5a", "messageHeadline": "add missing link to openvino (#10737)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f9b6eef05f58c5780ca31ed76a55aa5df1f488b0", "messageHeadline": "orttraining packaging pipeline for rocm 5.0.1 (#10725)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 249, "deletions": 3}}, {"node": {"oid": "7ab0c607b45b84141858702e7f026301d07dc8b8", "messageHeadline": "add qdq support of (un)squeeze and GlobalAveragePool (#10721)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 178, "deletions": 32}}, {"node": {"oid": "9ad95bf068e4e25bec79c8715440d9bed909b3c8", "messageHeadline": "Skip SetName test on inbox build (#10699)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "5d8c5409ab8b0be623cdf024ed2dbedcd847ffde", "messageHeadline": "POWER10: QGEMM optimization (#10642)", "author": {"name": "RajalakshmiSR", "email": "58744256+RajalakshmiSR@users.noreply.github.com", "user": {"login": "RajalakshmiSR"}}, "additions": 1205, "deletions": 0}}, {"node": {"oid": "e5c6dc1fc878857af02b80da8aba17b64f2f6299", "messageHeadline": "Add ability to save calibration augmented models through external dat\u2026", "author": {"name": "Funtowicz Morgan", "email": "mfuntowicz@users.noreply.github.com", "user": {"login": "mfuntowicz"}}, "additions": 50, "deletions": 20}}, {"node": {"oid": "62cc9815994eb5664f45ca603a124b06e12e54d1", "messageHeadline": "[TVM EP] support of TVM Virtual Machine (#10341)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 250, "deletions": 128}}, {"node": {"oid": "a7f6442c45ede16e92205e9d4c7068819eef37ae", "messageHeadline": "[js] release pipeline for web and react native (#10656)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 177, "deletions": 108}}, {"node": {"oid": "9e7d7a9e97c979c3b32fac4d54324e3f9c6b1d40", "messageHeadline": "Convert ConvActivationFusion transformer to a selector action transfo\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 671, "deletions": 363}}, {"node": {"oid": "fa9090f2597b1bba6db8fe5c85737f50f9dd0ef8", "messageHeadline": "check gpt-2 graph in converting beam search (#10712)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 46, "deletions": 44}}, {"node": {"oid": "d07a2377b10add786e9ed10fed97918af96004e9", "messageHeadline": "Fix race condition in CUDA, ROCm, and TensorRT EP GetKernelRegistry()\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 904, "deletions": 868}}, {"node": {"oid": "2fb2dae42f3907257c3da8018b29bf8801ae065a", "messageHeadline": "Print tensor snippet in dumping node Inputs/Outputs to StdOut (#10707)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 42, "deletions": 68}}, {"node": {"oid": "a7738b52c5fbeb6f29e33da2421fb3e0ad7dfd7e", "messageHeadline": "Add microbench to benchmark single operators. (#10678)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 129, "deletions": 0}}, {"node": {"oid": "19464614e7a71c42016dcf0c5bc3b9e7880975db", "messageHeadline": "[NNAPI QDQ] Add QDQ Concat  (#10666)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 212, "deletions": 58}}, {"node": {"oid": "6448ca64e6234c1a1b1bcc81da9850fd2823a179", "messageHeadline": "Fix reshape allowzero with unknowndim (#10665)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 21, "deletions": 5}}, {"node": {"oid": "f652f70d917f5a47a7bc7d5d06a41df575fd0611", "messageHeadline": "set qdq as the default static quantization format (#10684)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 86, "deletions": 52}}, {"node": {"oid": "f1b6f0becdd4aa45596d34e7461b2ab4caf770d9", "messageHeadline": "Update nuget icon (#10672)", "author": {"name": "Yi Zhang", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 7, "deletions": 5}}, {"node": {"oid": "c1cf16ed5d078b69db865c633e446a6038d22cea", "messageHeadline": "Conv node bug, cached state was incoherent (#10041)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "f4b2d3af2b6411d8b85f5c99b65d083185282418", "messageHeadline": "Upgrade emsdk to 3.1.3 (#10577)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 18, "deletions": 4}}, {"node": {"oid": "c51b500ca73dd8b9b8ec332735ac609e8dc6269d", "messageHeadline": "replace std::numeric_limits<T> by cub::FpLimits<T> (#10703)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "9a22b5d253be823b3a2532439476eb99b826e97d", "messageHeadline": "Strided Tensor Support for Eager Mode (#10578)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 542, "deletions": 176}}, {"node": {"oid": "f856608599dd1bc54f3f803c1e64fbf90cee765d", "messageHeadline": "[java] Changes OrtEnvironment so it can't be closed by users (#10670)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 643, "deletions": 668}}, {"node": {"oid": "e23a22451880a1ba1f46b2af6a71bb4a1079108b", "messageHeadline": "Fix CUDA 10.2 compile error due to inlined_containers.h inclusion (#1\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 12, "deletions": 11}}, {"node": {"oid": "3243c9579f9f120904d0664e89925a31e9d68139", "messageHeadline": "Fix VLOG?_DEFAULT macros usability. (#10568)", "author": {"name": "cloudhan", "email": "cloudhan@outlook.com", "user": {"login": "cloudhan"}}, "additions": 38, "deletions": 5}}, {"node": {"oid": "d1b2fb15addaa369660d617c6d39595c00e2a963", "messageHeadline": "Avoid clang-tidy crashing due to readability-static-accessed-through-\u2026", "author": {"name": "cloudhan", "email": "cloudhan@outlook.com", "user": {"login": "cloudhan"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d2d22f21955e79c072edbaa2ac1b8c1e2dba937f", "messageHeadline": "Add support of generating symmetric/asymmetric tensor's range for cal\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 99, "deletions": 23}}, {"node": {"oid": "ffde44cd09e1fdc1ff84c41bcca96ed4469a2330", "messageHeadline": "[iOS Packaging] Add full ORT build iOS package. (#10626)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 402, "deletions": 167}}, {"node": {"oid": "1f6d8248dadef0559e34730a3619f13ec4d30e6e", "messageHeadline": "Add optional optimizer to remove leftover Q->DQ pairs after all other\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 287, "deletions": 6}}, {"node": {"oid": "e788cc2a2370946977b91b2456166d466df4d14a", "messageHeadline": "Convert com.microsoft::ATen into org.pytorch.aten::ATen onnx op (#10060)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 115, "deletions": 111}}, {"node": {"oid": "e47434ea1281d83e54ea99ef053223aae874ec91", "messageHeadline": "[java] Adding the graph description to the exposed model metadata. (#\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 67, "deletions": 6}}, {"node": {"oid": "037f08f1ff445c39a33e4a03be88f744613593da", "messageHeadline": "Fix unsqueeze for opset 13 for ReduceMean Grad (#10668)", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 16, "deletions": 6}}, {"node": {"oid": "eb116595d4c4b28eedc009eae98e2bba777e0a1b", "messageHeadline": "Add ability to customize ORT_CXX_API_THROW (#10688)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "240f31ef6eb93d3ce672c1c6ec84cf92fdafaed1", "messageHeadline": "fix softplus (#10576)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 19, "deletions": 19}}, {"node": {"oid": "f2ca43fe0d6ab1156bb43128e76c283bd21e46c5", "messageHeadline": "Enable CoreML in the macos package (#10675)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 23, "deletions": 13}}, {"node": {"oid": "b30e0e2283ad814ca82b7eeceaf56579234ad962", "messageHeadline": "Remove inline_containers include from tensor_shape (#10682)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 196, "deletions": 105}}, {"node": {"oid": "81831201a8656059587518998653ca855c7f569e", "messageHeadline": "Change C# tests to use C# 5.0 (#10686)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "5fbfca3d58b88d0da29c6b206c36980052a4392c", "messageHeadline": "Add Experimental API for setting model name (#10518)", "author": {"name": "Numfor Tiapo", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 87, "deletions": 0}}, {"node": {"oid": "36c3271546b2d960b5e2511154b28b12c6aa4fb1", "messageHeadline": "BeamSearch op cuda (#10556)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3476, "deletions": 1225}}, {"node": {"oid": "957dccb37929b8e76d4c39fa7b2868bdc30c355e", "messageHeadline": "Fix compile (#10667)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "12c44bfc4eef5f9482a8f3d091c51ed52f2907f9", "messageHeadline": "fix bug: getting current cpu core type (#10630)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 107, "deletions": 71}}, {"node": {"oid": "617474e29884f7b20950f8731a8e33a9975bb45a", "messageHeadline": "Stop gradient edges for aten::argmax (#10650)", "author": {"name": "David Fan", "email": "30608893+jiafatom@users.noreply.github.com", "user": {"login": "jiafatom"}}, "additions": 24, "deletions": 0}}, {"node": {"oid": "2679711bee9717f8ab77e20c163ddf2fe3a3f48d", "messageHeadline": "Refactor transformers and other code to reduce memory allocation call\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1320, "deletions": 885}}, {"node": {"oid": "0b19a03361978f9576dd9d91629750b3ac12a4d4", "messageHeadline": "Fix the debug dump of tensor values to output int8 and uint8 values c\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 142, "deletions": 136}}, {"node": {"oid": "b993de9dd406b47c442734937be6e09914b61e01", "messageHeadline": "add reshape handler (#10627)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 237, "deletions": 49}}, {"node": {"oid": "56be66a0abb63de02ea71a4e2718baf21e8d9fc6", "messageHeadline": "Update c-api-cpu.yml: change nuget linux arm64 RID", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7660eeef3eff9a58b4d807dd951a34bafcb03ac0", "messageHeadline": "fix ortmodule's output device info when it runs on ort device (#10616)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 22, "deletions": 11}}, {"node": {"oid": "446258fa2864607f733ce6ecca130f38054a8039", "messageHeadline": "fix bug: quantize output of activation op(Relu, Clip) (#10649)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 87, "deletions": 1}}, {"node": {"oid": "7dc7529ec8e4384bddcd090f0c14c42c98cf37a1", "messageHeadline": "[TVM EP] Integrate tests for TVM EP into public onnxruntime CI (#10505)", "author": {"name": "Alexey Gladyshev", "email": "wotpricol@mail.ru", "user": {"login": "KJlaccHoeUM9l"}}, "additions": 220, "deletions": 85}}, {"node": {"oid": "ecf064f135d8e60b62f06d6010c335626c6eef44", "messageHeadline": "Exclude pdb from nuspec unless it's the winml package. (#10638)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "e076f3a125c27672407a7205bf36a47330e29ba2", "messageHeadline": "Fix incorrect target name", "author": {"name": "Thomas Rondahl", "email": "thoron@users.noreply.github.com", "user": {"login": "thoron"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "573e440d351c18fb1b575177ac13b0c05198e1b1", "messageHeadline": "Fix no ARM64 natives for Linux nuget", "author": {"name": "Thomas Rondahl", "email": "thoron@users.noreply.github.com", "user": {"login": "thoron"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bd08f11a5895194e05657cedc5b4e0b1e594a034", "messageHeadline": "Upsample support NHWC (#10554)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 342, "deletions": 275}}, {"node": {"oid": "e0d1d6906a04d69e90c06277c32ccff098077006", "messageHeadline": "Merge two helpers involving the kernel def hashes into one file (#10609)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 101, "deletions": 91}}, {"node": {"oid": "ea7f773a6e1244bab32ba791a3c855e4323d3135", "messageHeadline": "Merge pull request #10619 from microsoft/user/dwayner/DmlDev20220221", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 952, "deletions": 578}}, {"node": {"oid": "9ba2d9379ff83e0836037988228094d2fea36301", "messageHeadline": "[ROCm] Code sync from CUDA (#10631)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 21, "deletions": 4}}, {"node": {"oid": "8cfa4b1c175fa48ed13216b21af12ac3d71567e7", "messageHeadline": "Fix build errors due to changes in warnings that VS 2022 17.1 produce\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "7de86d39d3bc17b552b0fd1cd585e9755f66311b", "messageHeadline": "Build error int to bool", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d6a8cba27302bcd494040f32511568f81d70bace", "messageHeadline": "[NNAPI QDQ] Add nnapi qdq softmax op support (#10591)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 152, "deletions": 24}}, {"node": {"oid": "4d3cd2f68526e39d6ae75f68d68f9528b5145ec6", "messageHeadline": "Add helper for optimizing a QDQ format model for usage with ORT. (#10\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 164, "deletions": 0}}, {"node": {"oid": "4a79ed62b4bc08ecf24c339712e6865004466190", "messageHeadline": "Remove extra version of a function in dnnl (#10599)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 0, "deletions": 4}}, {"node": {"oid": "742694f6796a3ce4426881f51eb3fddd38abb08e", "messageHeadline": "[python] [orttraining] Add utility to export a graph to compute gradi\u2026", "author": {"name": "Justin D. Harris", "email": "juharris@users.noreply.github.com", "user": {"login": "juharris"}}, "additions": 292, "deletions": 4}}, {"node": {"oid": "6f0640a57f29ec0705a35b6054e97fd60f3d564b", "messageHeadline": "Optimize ReduceSum, ReduceMean, ReduceMin, ReduceMax (#10280)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 520, "deletions": 56}}, {"node": {"oid": "df841ee87d0e3c9b814dabaadb4f0490e374a3e6", "messageHeadline": "Fix incorrect type constraint registration for operator kernels. (#10\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1362, "deletions": 1167}}, {"node": {"oid": "893ee65e54db599e7b8486dd78999402739e53ef", "messageHeadline": "[js/web] fix lint error when run without ort-web TS types (#10429)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "6db6ee5710be7d247e7eace007dbfca1917a4cf8", "messageHeadline": "Merged PR 6973543: ORT DML EP Opset 13 more complete", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 952, "deletions": 578}}, {"node": {"oid": "1af4c170ef8dc869ba732043f6f4a38b07338e05", "messageHeadline": "[js/react_native] publish onnxruntime-common npm package as web and n\u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 17, "deletions": 4}}, {"node": {"oid": "e056fbaa51e5e52793b14dc9bc0a4fe117e3c711", "messageHeadline": "Add restrictions for hybrid cpus for thread pool task distribution (#\u2026", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 25, "deletions": 10}}, {"node": {"oid": "2fa333443a6639708e664b51c279c9e10c99431b", "messageHeadline": "Add telemetry for device kind (#10431)", "author": {"name": "Jingqiao Fu", "email": "jingf@microsoft.com", "user": null}, "additions": 20, "deletions": 0}}, {"node": {"oid": "2ca9566994d019e5e310212c9c8912147af300e0", "messageHeadline": "Add range of helpers for making usage of ORT Mobile easier. (#10458)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1958, "deletions": 100}}, {"node": {"oid": "fad590a05908406d50bd55e8f653d61c9ebdc524", "messageHeadline": "Enhance TRT EP unit tests (#10493)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 415, "deletions": 205}}, {"node": {"oid": "edbc844032ec7abc429f62df12016e68093194ba", "messageHeadline": "Fix misspelling in python documentation (#10588)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fd16085cea3d466a3e71a7cfb47373b616f796a1", "messageHeadline": "Zhanyao/attention (#10545)", "author": {"name": "zhangyaobit", "email": "1034716+zhangyaobit@users.noreply.github.com", "user": {"login": "zhangyaobit"}}, "additions": 773, "deletions": 7}}, {"node": {"oid": "8d06e5a9df3024dbaadf4e221812d2fd46faabcd", "messageHeadline": "Add openvino base image option (#10581)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 23, "deletions": 3}}, {"node": {"oid": "ccd7a2d8407c37c264258cb79725cd2cb6b31328", "messageHeadline": "Fix build failure when using clang compiler", "author": {"name": "Pallavi Deshmukh", "email": "pdeshmuk@microsoft.com", "user": {"login": "pallavides"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "09ac7595fcd327f48038dff344bd09792a3a5951", "messageHeadline": "update (#10573)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "4f76c386862bfe6dcba5a427969f0071b95a9ab9", "messageHeadline": "Revert \"Reduce max gradient (#9859)\" (#10574)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 3, "deletions": 81}}, {"node": {"oid": "e71d77e9740d5ae04b5eb0d29e7811362150e4f5", "messageHeadline": "[ROCm] Fixing build for CIs (#10558)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 12, "deletions": 14}}, {"node": {"oid": "3be3db51804fe54a44539742b33bc1dabef39574", "messageHeadline": "Use IntPtr instead of int conversion for pointer in Memory.Pin() (#10\u2026", "author": {"name": "Lingfeng Wu", "email": "handtalk@gmail.com", "user": {"login": "fillmore"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d198fbc4d5e6e60ddf80bc2dcfbf7c3cf9cb62d4", "messageHeadline": "Add a script for randomizing onnx weights (#10551)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 90, "deletions": 0}}, {"node": {"oid": "7443edb0bfc0e147b660b8f43e0e73df8117a98a", "messageHeadline": "Reduce max gradient (#9859)", "author": {"name": "Anh Nguyen", "email": "94985387+anhnguyen7198@users.noreply.github.com", "user": {"login": "anhnguyen7198"}}, "additions": 81, "deletions": 3}}, {"node": {"oid": "f436d3437e85fa0c28a6264dd361958ae4d63e08", "messageHeadline": "Add layout transformer for NNAPI (#10371)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 1107, "deletions": 710}}, {"node": {"oid": "ceb1e2b1a6eb4a14d37b202c8f25e666f5c95ac7", "messageHeadline": "[ROCm] Bugfix of BFloat16-float conversion and Add FastGelu Kernel fo\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 86, "deletions": 37}}, {"node": {"oid": "f22cd3af5d0c2026fa0ad9861170e2d05b96b70d", "messageHeadline": "Leqiao/add selectable pipeline (#10560)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 625, "deletions": 2}}, {"node": {"oid": "05d6805830ddeff6c4229ade7d1ae07c97d97f9b", "messageHeadline": "clean up quantization of QAT model (#10549)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 12, "deletions": 60}}, {"node": {"oid": "8e47bb9a4a3a630a3731ce19edbde6a5546b2a50", "messageHeadline": "[NNAPI QDQ] Add QDQReshape op support (#10533)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 114, "deletions": 28}}, {"node": {"oid": "0c3e88944d789a6779f3722eb0cf67710b3dab3f", "messageHeadline": "Fix create ort value hardcoded memory info to CPU (#10510)", "author": {"name": "Anh Nguyen", "email": "94985387+anhnguyen7198@users.noreply.github.com", "user": {"login": "anhnguyen7198"}}, "additions": 32, "deletions": 12}}, {"node": {"oid": "1cdc23aba419cab17aa9cc5faedfdd1bef8ab023", "messageHeadline": "[TVM EP] Rename Standalone TVM (STVM) Execution Provider to TVM EP (#\u2026", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 611, "deletions": 595}}, {"node": {"oid": "d3f7459263eb14f21ec8aa46d691a8cf1f07e0f0", "messageHeadline": "fix CI build (#10553)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "58f80c16ff50a10d59f911059388a860d4571fa9", "messageHeadline": "Create branch according to cpu core uarch (#10521)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 102, "deletions": 2}}, {"node": {"oid": "3199074ac786d655d4390b4dcd007a445534633e", "messageHeadline": "Update QDQ propagation transformer to insert QDQ nodes (#10487)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 646, "deletions": 291}}, {"node": {"oid": "7691e7ed12f9722c8f984ca056c3aef1c11ec900", "messageHeadline": "Introduce load balancing dataset samplers (#10163)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 575, "deletions": 2}}, {"node": {"oid": "270dec73277e4dc63ee1cdcee9278524fe577da0", "messageHeadline": "Return a Status instead of throw an exception in GetAttrs (#10534)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 29, "deletions": 28}}, {"node": {"oid": "3f37609994c12d3eb6328c80843cb80a541aeddf", "messageHeadline": "Remove unneeded code in UpsampleBilinear (#10544)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "bfb20b315d9763f4113fdb1b15444626d66874e6", "messageHeadline": "Bump karma from 6.3.2 to 6.3.14 in /js/web", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 212, "deletions": 110}}, {"node": {"oid": "5cfde7af29549b407daaaa1c228fb144a5375981", "messageHeadline": "[NNAPI QDQ] Add QDQTranspose op support (#10495)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 86, "deletions": 22}}, {"node": {"oid": "318d31ea127352dc451c5fb4c6e508daa49290ad", "messageHeadline": "Fix C# pipeline build error (#10524)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4e2a9740907789c6dc537117fee76fce268413c7", "messageHeadline": "[ROCm] UTs and code clean up (#10511)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 103, "deletions": 93}}, {"node": {"oid": "2002a965943a01dfb3f50c636cceedd8b14553b9", "messageHeadline": "The transformer of memcpy is needed for ROCm EP and MIGraphX EP when \u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "f92e47e95b13a240e37caf7b36577983544f98fc", "messageHeadline": "Remove onnxruntime_util dependency on onnxruntime_framework (#10512)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 88, "deletions": 58}}, {"node": {"oid": "a27aabad34ae814af6611550057f16d6f0096810", "messageHeadline": "Fix fomatting. (#10520)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "3185680b6c8f5be6aafd2001f2d0ce74299a0569", "messageHeadline": "Add NHWC CONV contrib op (#10506)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 168, "deletions": 1}}, {"node": {"oid": "eba730500fd7d10fafe057fad364f9543346ad7a", "messageHeadline": "Remove file-scope non-constant static variables to support multiple i\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 550, "deletions": 514}}, {"node": {"oid": "4d6d4dfb9de610fbb09e23b36379473b4005a99d", "messageHeadline": "Add TRT ep perf benchmark (#10470)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 105, "deletions": 53}}, {"node": {"oid": "dd33ce0fdc6eab593247b613bfce23b4f2b5b209", "messageHeadline": "[js/react_native] Create ONNX Runtime React Native pipeline (#10474)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 2089, "deletions": 312}}, {"node": {"oid": "6f3ade55ecd2b989612c1680b87ad9b6c557d1c9", "messageHeadline": "Move QAttention/QEmbedLayerNormalization op defs to quantization_defs\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 317, "deletions": 289}}, {"node": {"oid": "c9fbd0b15af9f2b595e59632c146f997790ed25c", "messageHeadline": "Optimize cuComputePartGradGammaBeta kernel for MI100 (#10475)", "author": {"name": "Hubert Lu", "email": "55214931+hubertlu-tw@users.noreply.github.com", "user": {"login": "hubertlu-tw"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "7a2bf3c24c7ba74e04b8bfcd0d1ece969c35b562", "messageHeadline": "Reorganize contrib op schemas (#10494)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3068, "deletions": 2962}}, {"node": {"oid": "399ffc970068724407e8225e20907f69ea917603", "messageHeadline": "Fix Windows GPU CI (#10499)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "e4dc4e4d3c79a7346dbd78e9badc89dfac384bee", "messageHeadline": "[NNAPI QDQ] AddQDQAdd/Mul, update to NNAPI QDQ handling, update some \u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 710, "deletions": 745}}, {"node": {"oid": "655f490c9564c27543d53121c20a157cb2aba76b", "messageHeadline": "Remove BFloat16 Specialized Code for ReduceSum (#10476)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 34, "deletions": 214}}, {"node": {"oid": "4388eaed1bd6a4afb2744754dac4b5288295d941", "messageHeadline": "Merged PR 6937750: Restore history to dmldev. Merge without squash", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "b14944f9f85d3f1b062b7caf84b9f2a3b76a75a1", "messageHeadline": "Merge commit 'b02f4ece5e4f48f5d303d6be0170c03d60b24efb' into user/ryl\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "7e5d68eea664e362c99c298e8747a8b7314dab8b", "messageHeadline": "gradient and test (#10455)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 75, "deletions": 0}}, {"node": {"oid": "435e14d60aa7fa7e53f25bc73b65d41642b951e4", "messageHeadline": "[ROCm] BFloat16 support (#10465)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 471, "deletions": 65}}, {"node": {"oid": "c696da36c70dc7befc74377d10cce7a29486e750", "messageHeadline": "fix unit test of quant gemm (#10469)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0f5d0a091a37db48593905c33a834597338288e8", "messageHeadline": "Make user capable of adding new field in OrtTensorRTProviderOptionsV2\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 84, "deletions": 9}}, {"node": {"oid": "927f1f18c9db4614ec5cab021c1041bb55766694", "messageHeadline": "[NNAPI QDQ] Add QDQ AveragePool op support (#10464)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 73, "deletions": 47}}, {"node": {"oid": "d0ab881d070756d50d8b75fa03e80d78f9b0a0cc", "messageHeadline": "Contrib ops for TRT plugins: EfficientNMS and Pyramid ROI Align (#9486)", "author": {"name": "wraveane", "email": "75912517+wraveane@users.noreply.github.com", "user": {"login": "wraveane"}}, "additions": 163, "deletions": 0}}, {"node": {"oid": "6fd7ba5b7e7113cc959c7b36622a621dc198a3da", "messageHeadline": "Merged PR 6917440: ONNX Runtime update from GitHub master", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 23837, "deletions": 6051}}, {"node": {"oid": "0d09dd5d20c3a5133e6e6ae1218cdc45ccaf0358", "messageHeadline": "Support fusion for TNLR based model (#10432)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 190, "deletions": 8}}, {"node": {"oid": "4f13c8ac3980239d29f834cf6bf9688b96799c21", "messageHeadline": "Update orttraining-linux-ci-pipeline.yml (#10462)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "6bbf016dc4561a45191667087a9326516988a4ea", "messageHeadline": "cmake: disable 'attributes' error to fix the build with GCC < 9.x", "author": {"name": "Maxiwell S. Garcia", "email": "maxiwell@linux.ibm.com", "user": {"login": "maxiwell"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "bb09acffed8323ea13664bbff96c4914b875b477", "messageHeadline": "Transformer model CUDA EP align with CPU on corner case (#9889)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 59, "deletions": 15}}, {"node": {"oid": "63198a6566e8b3703f84477db0ec7a313cc69dd2", "messageHeadline": "[ROCm] BFloat16 support  (#10447)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 342, "deletions": 11}}, {"node": {"oid": "239c6ad3f021ff7cc2e6247eb074bd4208dc11e2", "messageHeadline": "Support specifying an execution provider in benchmark script (#10453)", "author": {"name": "zhangyaobit", "email": "yaozhang@google.com", "user": null}, "additions": 79, "deletions": 22}}, {"node": {"oid": "a405658370cc05989209c6608ff9f591799c84ae", "messageHeadline": "Fuse Clip->Q to Q (#10434)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 280, "deletions": 65}}, {"node": {"oid": "97b8f6f394ae02c73ed775f456fd85639c91ced1", "messageHeadline": "Add logic to NNAPI EP to exclude pre-processing involving dynamic sha\u2026", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 94, "deletions": 37}}, {"node": {"oid": "6076a262dccf9b8dc024e73d7401e1183bdb234d", "messageHeadline": "upgrade react-native packages to latest (#10454)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 2754, "deletions": 4643}}, {"node": {"oid": "ad9d2e2e891714e0911ccc3fa8b70f42025b4d56", "messageHeadline": "Prefix match in first iteration of beam search OP (#10231)", "author": {"name": "Viswanath Boga", "email": "44417868+viboga@users.noreply.github.com", "user": {"login": "viboga"}}, "additions": 134, "deletions": 12}}, {"node": {"oid": "1aa078969163c9cb1c70f07a65362c2e781fbf12", "messageHeadline": "add qdq support for QGemm (#10414)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 579, "deletions": 68}}, {"node": {"oid": "73183616451655a0912b7d5068859036f804d035", "messageHeadline": "[NNAPI QDQ] Add QDQ Resize support (#10442)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 145, "deletions": 35}}, {"node": {"oid": "91b8ad5ee73df834782b19093da7038f6e0ab6e1", "messageHeadline": "Allow users to bind arbitrary memory using raw pointers (#10428)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 274, "deletions": 49}}, {"node": {"oid": "3c967601921093ebffcdbed0a1baf1dc9f8b8f3d", "messageHeadline": "support rocm/migraphx EP in perftest tool (#10449)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "062129a5c45cec01de0b5c92371993d3ed30412f", "messageHeadline": "Update rocm_ep and migraphx_ep to rocm4.5.2 and fix dockerfiles to bu\u2026", "author": {"name": "Shucai Xiao", "email": "shucai@gmail.com", "user": {"login": "scxiao"}}, "additions": 54, "deletions": 43}}, {"node": {"oid": "a1d9a71b8ba9426e3448ab02aa00de82403bd113", "messageHeadline": "Improve Perf System (#10404)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 60, "deletions": 38}}, {"node": {"oid": "a7c67860a532b3fc8e579d9b5407be432f07cd35", "messageHeadline": "Reduce test time for TensorRT EP CI (#10408)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 51, "deletions": 6}}, {"node": {"oid": "ef7b4dc05cae1084546ff7caba048a2f908ac1d8", "messageHeadline": "Add test quantization of ArgMax for TensorRT (#10325)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "68262cce86bff8805b0bae9b26030845e7e87d38", "messageHeadline": "[NNAPI QDQ] Add QDQ Conv support (#10418)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 117, "deletions": 101}}, {"node": {"oid": "c43c1691ad6945f1e20d30408c70a04c227a35eb", "messageHeadline": "Enable transpose optimizer in minimal extended build (#10349)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 318, "deletions": 222}}, {"node": {"oid": "baa17679221807ea332a7134c0d10b81adf4009e", "messageHeadline": "Allow for an optional subgraph input to have no type info. (#10379)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 26, "deletions": 21}}, {"node": {"oid": "85cbe8367e8576cde03435666c02964aeceedaa6", "messageHeadline": "[ROCm] BFloat16 support (#10416)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 177, "deletions": 37}}, {"node": {"oid": "b02f4ece5e4f48f5d303d6be0170c03d60b24efb", "messageHeadline": "Remove cbegin and cend calls which do not exist in std::span or gsl::\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "5f0ba31890249b4917e20fecef46144509b6fc10", "messageHeadline": "Remove coremltools submodule *security vulnerability*  and copy the c\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 8912, "deletions": 84}}, {"node": {"oid": "c4f1dfcfaa1d68567606d2ad919330e2daecd742", "messageHeadline": "Cfu s8s8 (#10413)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 3701, "deletions": 222}}, {"node": {"oid": "1a2925acce7abf77575d8ffd03d51010bf772e59", "messageHeadline": "Add sympy package as a dependency (#10406)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2dd5e75ba87936f2681557286b6fc83b95451cf4", "messageHeadline": "Incorrect output after GPU to GPU inference via VideoFrame and Gray8 \u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "feae842a7c9330733195e490fb6b09663c21b630", "messageHeadline": "Update pytorch-lightning (#10421)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b14da94fc1c77bd9b78860d3c8a5d4f0fea3ef71", "messageHeadline": "Exclude CETCOMPAT from Windows ARM build (#10417)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ce081fe6558d97b27d62ce8d3d69a42833e9f8df", "messageHeadline": "Fix TopK with NAN on Cuda (#10314)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "ff2057a817c50d52e0bceeb8cd3ea7b19403eb27", "messageHeadline": "Add sample qdq unit test case for nnapi ep qdq integration (#10358)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 134, "deletions": 71}}, {"node": {"oid": "0e951d7d6b3da1b1389738302bac778bb59fbeda", "messageHeadline": "Add some more documentation for the C/C++ API tensor creation functio\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 44, "deletions": 18}}, {"node": {"oid": "481b96d32aecc5fceabc9e9401604d2d31212ef7", "messageHeadline": "STVM, NUPHAR, remove tvm from submodules list, checks pointers are no\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 332, "deletions": 101}}, {"node": {"oid": "ec4362f8f326e076498ad2fde1aa44929b586f25", "messageHeadline": "Enable more static analysis warnings and enable the analyzer for trai\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 416, "deletions": 204}}, {"node": {"oid": "66acf50488c210ee89f490aea40d17256d14e1ea", "messageHeadline": "Document C/C++ API documentation version info conventions. (#10396)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 20, "deletions": 1}}, {"node": {"oid": "3367ddc5ba1f71f8416720aebb302bba19e75b79", "messageHeadline": "Add abseil cgmanifest declaration. Update coding standards. (#10374)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 162, "deletions": 44}}, {"node": {"oid": "4d305282da4a4c0bd20447ff633db852d65f3d23", "messageHeadline": "[ROCm] Enable BFloat16 for Gemm and MatMul Op (#10398)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 176, "deletions": 7}}, {"node": {"oid": "5f49f40fa56e3098ef8b3839212939c325fb5cdc", "messageHeadline": "Bump log4js from 6.3.0 to 6.4.0 in /js/web", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 53, "deletions": 65}}, {"node": {"oid": "27a4af6074da28d3d53315f577a351203412cb83", "messageHeadline": "Fix some BinSkim defects (#10400)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "c6ef465011bd395bd0a1a16b820cabfa585768b2", "messageHeadline": "minor fix in node unit change (#10405)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 13, "deletions": 12}}, {"node": {"oid": "ea9c8a7cdcd7fd4e13bd0920f883a3c5f1f76918", "messageHeadline": "support MIGraphXEP to work with ROCMEP for inference on AMD GPU (#10368)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 29, "deletions": 81}}, {"node": {"oid": "389d2db1ce8e080382cdd1411921bfd9e6a0221b", "messageHeadline": "Make model tests name clear (#10220)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 40, "deletions": 1}}, {"node": {"oid": "847801f5bef3773d04fcdb0cfba1daa854bbb7d4", "messageHeadline": "[wasm] update emscripten v2.0.34 (#10391)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 16, "deletions": 3}}, {"node": {"oid": "cf13b9dd5efe9f4821f27a2078ad6faf6959600a", "messageHeadline": "Symbolic export for numpy_T  (#10390)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 47, "deletions": 0}}, {"node": {"oid": "a27503ebe4fd0ccb9d3454c78f2fe0eca9919c5e", "messageHeadline": "use strict mode (#10397)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "5576e3553dd2a336389bf1b08a1f0cbebda547d7", "messageHeadline": "Remove python 3.6 from our python packaging pipeline (#10395)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 36}}, {"node": {"oid": "4af116649c8f5f6e725ce8b314b7f8e38007f236", "messageHeadline": " [QDQ] Hookup NNAPI GetCapability/Compile with shared QDQ selectors (\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 448, "deletions": 148}}, {"node": {"oid": "9aa51379c93184a94171bc7902725d327e8722c5", "messageHeadline": "[eager mode]: add configuration for ort virtual device count (#10346)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 16, "deletions": 18}}, {"node": {"oid": "5eafbb50f9ad2b41885fb61a6278708b7264ed2a", "messageHeadline": "Fix possible null pointer dereference. (#10373)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e1012a8662585b730ba1973450f1c7e0911735d5", "messageHeadline": "Added OnRunEnd and Sync method in ExecutionProvider (#10362)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 24, "deletions": 0}}, {"node": {"oid": "df16c605e8aa287c7936e5fe80fdc337e744804b", "messageHeadline": "Add \"available since\" message for C API additions since v1.10.0. (#10\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 42, "deletions": 26}}, {"node": {"oid": "a0fe4a7c1c156d43840af42111a559dbbd9753ed", "messageHeadline": "[TVM EP] Improved usability of TVM EP (#10241)", "author": {"name": "Alexey Gladyshev", "email": "wotpricol@mail.ru", "user": {"login": "KJlaccHoeUM9l"}}, "additions": 30, "deletions": 0}}, {"node": {"oid": "6e95c0316d8375ac1926f77cde75d80bcdd453d6", "messageHeadline": "Builds onnxruntime + eager mode with the same value for _GLIBCXX_USE_\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "790c3be7e9117432df0f93128c5a2271fd6f1b40", "messageHeadline": "Fix Reshape issue when shape size is -1 (#10356)", "author": {"name": "pallavides", "email": "pallavides@users.noreply.github.com", "user": {"login": "pallavides"}}, "additions": 14, "deletions": 6}}, {"node": {"oid": "4b87d2c172bc77ca67a674009e48e3eb30cc155a", "messageHeadline": "Fix dockerfiles/Dockerfile.arm32v7 build. (#10360)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 42, "deletions": 4}}, {"node": {"oid": "df0c819850d8426706e853b629af3803ad5cc469", "messageHeadline": "fix compilation error due to symantic conflict with another PR (#10370)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "2afce4830c9ff48374ca11cf00b09afcc1ff2124", "messageHeadline": "Symmetric QGEMM  (#10289)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 3148, "deletions": 103}}, {"node": {"oid": "7e092a7e3f8f9f590de0ac5548035e8fa4d512f8", "messageHeadline": "Reduce number of memory allocations based on a customer profiling cas\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2095, "deletions": 1595}}, {"node": {"oid": "5df15c56444fe1b5f770eaca38d8031b5ffe151e", "messageHeadline": "additional options of NNAPI for ORT_PERF_TOOL (#10351)", "author": {"name": "wejoncy", "email": "247153481@qq.com", "user": {"login": "wejoncy"}}, "additions": 27, "deletions": 3}}, {"node": {"oid": "3dfadf9031e6a80f487719e6603c2cc2f75fe65b", "messageHeadline": "[FIX] Add condition in amd ci pipeline yaml to stop test in time when\u2026", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 15, "deletions": 9}}, {"node": {"oid": "42db893607b3a4789cc2a13924269986a781ca9c", "messageHeadline": "Add ThresholdedRelu to ROCm EP. (#9480)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "6876641c1e6f270398a2345e2245b20a4e04454d", "messageHeadline": "Pin version of post to dashboard scripts' dependencies and update the\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 18, "deletions": 17}}, {"node": {"oid": "bfabef081dfb16226b056f33440d98ee7861f9d5", "messageHeadline": "Remove unused pipeline orttraining-linux-gpu-perf-test-ci-pipeline.ym\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 337}}, {"node": {"oid": "141606534c8e9f3089437b2300a7a715ec8b7d54", "messageHeadline": "Add support for FusedAdam to be mathematically equivalent to pytorch/\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 89, "deletions": 20}}, {"node": {"oid": "13e277525cff3b9e9d8a0b712bf191893ec9a114", "messageHeadline": "fix whitelist", "author": {"name": "Cheng Tang", "email": "chenta@microsoft.com@orttrainingdev9.d32nl1ml4oruzj4qz3bqlggovf.px.internal.cloudapp.net", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "eee627fde9da46e279154a15af581f621f2873fc", "messageHeadline": "Track Session Creation Time (#10281)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 122, "deletions": 31}}, {"node": {"oid": "d2b14249683d3e25994c2209027dbf5df9d98c16", "messageHeadline": "fix bugs in cpuid_info (#10334)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 64, "deletions": 69}}, {"node": {"oid": "2dcb69685e8b48a24d5a01e34e58bd0c0270dca9", "messageHeadline": "support type promotion in binary poerators in eager mode (#10285)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 163, "deletions": 32}}, {"node": {"oid": "c67594694c7d3cc8a27032318706e504e16c9c7b", "messageHeadline": "Add ability to set onnx opset version from json config (#10223)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 28, "deletions": 6}}, {"node": {"oid": "001cc539683d5e294d7b306d57e5d5bbb8422d73", "messageHeadline": "Fix CUDA10.2 Build Break for BFloat16 Change (#10331)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 45, "deletions": 66}}, {"node": {"oid": "4aa7cee0d8baf7485932b64370b4fc7dc1611874", "messageHeadline": "Abjindal/clean eager backend (#10055)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 50, "deletions": 12}}, {"node": {"oid": "90e2a4b936d0ab79806d5e1ecb87cc45b11537df", "messageHeadline": "Fix GH Issue 10305 by adding implicit inputs to consumer nodes map. (\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 258, "deletions": 183}}, {"node": {"oid": "a656c55a754f3758872b8864aa5640a7e2be6a48", "messageHeadline": "Add _force_exportable_set and pass debug_options (#10282)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "7b14c70cfe68b3a51b6e98a74b3ca89ed0493dae", "messageHeadline": "[ortmodule] Ensure contiguous tensor into forward pass (#10315)", "author": {"name": "David Fan", "email": "30608893+jiafatom@users.noreply.github.com", "user": {"login": "jiafatom"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "b038f4e56f90a810b47cf8176b84306a4c8bed57", "messageHeadline": "Add a build option to create a WebAssembly static library (#10184)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 254, "deletions": 64}}, {"node": {"oid": "03aab98adcc5a511722b1e4000ba935b7b5cee6a", "messageHeadline": "Merged PR 6857190: RI 1/19/2022", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 1058, "deletions": 849}}, {"node": {"oid": "810cd1275342811d09f5cfe05ca212136fbe597d", "messageHeadline": "Merge commit 'c12cafa524f254dc2ed1c80b22fa1821239fa137' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 1058, "deletions": 849}}, {"node": {"oid": "62eab67f796bf6ae693542cacb3f9fa5ac69829e", "messageHeadline": "Fuse DQ -> ArgMax into ArgMax (#10274)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 135, "deletions": 0}}, {"node": {"oid": "e27f2dc932f7c875c7f300cb5ebd4f12c62e6681", "messageHeadline": "int8/uint8 support for Argmax for opset 1, 11, 12 (#10296)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 36, "deletions": 4}}, {"node": {"oid": "712f4e403d9fa962e0104ad6c785bfdab6cc0007", "messageHeadline": "[js/common] upgrade marked@4.0.10 (Dependbot warning) (#10313)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1150, "deletions": 1170}}, {"node": {"oid": "c1c9fa18bfe257929dae7dc7447cfd672e232c73", "messageHeadline": "C#: Avoid inefficient DenseTensor ctor in ToTensor extensions (#10240)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 241, "deletions": 26}}, {"node": {"oid": "6ae22d562b0030190e1ee662118cbed49e9b7fdc", "messageHeadline": "[QDQ] Move NNAPI EP to use NodeUnitIODef for non-QDQ ops (#10237)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1217, "deletions": 1202}}, {"node": {"oid": "33dd2f8f5e517f09871975f64db974acee3376b2", "messageHeadline": "fix mac compilation error (#10268)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c12cafa524f254dc2ed1c80b22fa1821239fa137", "messageHeadline": "Optimize Transpose CUDA Kernel (#10230)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 46, "deletions": 13}}, {"node": {"oid": "a757bd7186deb855634b70c1a583d18e98cfa9ca", "messageHeadline": "Render summarized ort perf with tree map in browser (#10189)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 156, "deletions": 0}}, {"node": {"oid": "ab5fd42ed4a8499fc7ca13dcf1d67a44a299dd8c", "messageHeadline": "reset MIN for float/double (#10284)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 46, "deletions": 7}}, {"node": {"oid": "e365ad7f3abc6c99274e9481d0b509b2027e621d", "messageHeadline": "fix deadlock in model.train mode forward run only (#9960)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 54, "deletions": 17}}, {"node": {"oid": "6a7d3deb22bd78c2c243e1c8143000663127be5d", "messageHeadline": "Update pytorch-lightning (#10276)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "13ff79d1e4c042f919f34ed19794b0b3fbcd2c10", "messageHeadline": "Merged PR 6839685: Merge public onnxruntime", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 7705, "deletions": 7426}}, {"node": {"oid": "44e2db939711cf1c6108bee3088f69aef38553e4", "messageHeadline": "CUDA BFloat16 Refactor (#10085)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 707, "deletions": 764}}, {"node": {"oid": "e38e51ea8e4f04b11663b4261b163b7e39c38eee", "messageHeadline": "Improve iobinding, faster name search (#10005)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 33, "deletions": 35}}, {"node": {"oid": "3ea7fb0f9f8e4d40a0b0dd7f998eb16fbb379d7d", "messageHeadline": "fix mem leak (#10272)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "2a55bc2c21e6295b56f2744d803d76b213d99b2b", "messageHeadline": "Bump engine.io from 4.1.1 to 4.1.2 in /js/web", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "2affd6e71e08199eaa2c0b1148f94334fe9fdd43", "messageHeadline": "orttraining packaging and ci pipelines to use cuda 11.3 (#10252)", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9, "deletions": 8}}, {"node": {"oid": "6cb18fc686730ba4c4444c80d71c2357a43426f0", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into dmldev_temp", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 7705, "deletions": 7426}}, {"node": {"oid": "4b205eb2b302d6f8ca0aa19951e05a1059137e26", "messageHeadline": "Bump follow-redirects from 1.13.3 to 1.14.7 in /js/web (#10266)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 4895, "deletions": 4894}}, {"node": {"oid": "943a1aa2d66626517d28ed23dc4df3f523d4988f", "messageHeadline": "Bump follow-redirects from 1.14.5 to 1.14.7 in /js/node (#10265)", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 2457, "deletions": 2456}}, {"node": {"oid": "d43ef67d2b6ffc80823d8707862ee5d3443ac5cb", "messageHeadline": "Move binary size check to separate pipeline (#10254)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 153, "deletions": 67}}, {"node": {"oid": "3d9d8e20ccb753995e7a29e35f40572f129d10d2", "messageHeadline": "Bump numpy from 1.19.2 to 1.21.0 in /tools/ci_build", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8aad46c09b8eaa649e127f3a6f5a26092b18be17", "messageHeadline": "Merged PR 6836425: Merge RI without squash", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "a44d24d8b5b0c492cbfb0a2309dc20964ed463d8", "messageHeadline": "Merge remote-tracking branch 'origin/dmldev' into user/rylai/ri_01_12_22", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "c07e251cec3ba32e6cca1e8d895157a5c960bee6", "messageHeadline": "Merged PR 6835169: RI 12/9/21 - 01/12/22", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 33581, "deletions": 17886}}, {"node": {"oid": "499f1d5fd711e24cba7cda3ce30050fbe1894d42", "messageHeadline": "Quantization of Argmax (#10213)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 196, "deletions": 3}}, {"node": {"oid": "6f7389bc03b9f93e4b1988186e9283452daba320", "messageHeadline": "Merge commit '4af232df0ce89f0f0227ca8bd10ed8808d4398e5' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 33581, "deletions": 17886}}, {"node": {"oid": "98f85ae05b66785c0fcf5fdd5150fbe2e1a5f224", "messageHeadline": "Bump winrt version (#10243)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "aff96ce08180588196bed3d6df74c0099242e81c", "messageHeadline": "remove hardcoded type (#10251)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "4af232df0ce89f0f0227ca8bd10ed8808d4398e5", "messageHeadline": "Fix props file overwriting AdditionalIncludeDirectories (#10124)", "author": {"name": "CarlPoirier", "email": "carl.poirier.2@gmail.com", "user": {"login": "CarlPoirier"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a099bd454b8f4cb5098f2a45f757b1cc40e73869", "messageHeadline": "[QDQ] Add shared qdq selectors (#10178)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 256, "deletions": 0}}, {"node": {"oid": "79d2a0d18572397198553e17d046568877aa502b", "messageHeadline": "Dynamic cost model to mitigate high E2E perf variance  (#9833)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 172, "deletions": 60}}, {"node": {"oid": "fb4dea39e2c1635325ea4eb39c586d073daa9973", "messageHeadline": "Tolerate cpuinfo init failure (#10199)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 27, "deletions": 6}}, {"node": {"oid": "4048ed326c62f60ed7b1440f1e72dd2b7dd0b395", "messageHeadline": "Update EP Perf Pipeline (#10149)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 643, "deletions": 239}}, {"node": {"oid": "cb9b0275b65b56e127cba8d98dc451ee36f3a715", "messageHeadline": "Set default quantization weight type to int8 (#10209)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 70, "deletions": 52}}, {"node": {"oid": "ce103ace934bd397ec4e58e066fd625147b64c85", "messageHeadline": "Amdmigraphx fix build error (#9272)", "author": {"name": "Shucai Xiao", "email": "shucai@gmail.com", "user": {"login": "scxiao"}}, "additions": 1146, "deletions": 464}}, {"node": {"oid": "be9cc40aa51679daed51c88f9ce8ceecd5e6a800", "messageHeadline": "Make some parameters configurable for calibration (#10204)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 37, "deletions": 13}}, {"node": {"oid": "32ee379f50721f86d7ad4fdc93d30e277a6d02b5", "messageHeadline": "GraphPartitioner.cpp: fixed Merge function to handle duplicated parti\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5cd57bb726a2c3751cc4e8f93e2c5e8b29223cd8", "messageHeadline": "add load from buffer (#10162)", "author": {"name": "Jingqiao Fu", "email": "jingf@microsoft.com", "user": null}, "additions": 116, "deletions": 4}}, {"node": {"oid": "edd1a2cf61dd0c08bfc7a03e72a42e3e06d22014", "messageHeadline": "Add more Java test logging. (#10221)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "0f5e82c294fe719b1309399d7c5709c8351fb74d", "messageHeadline": "DirectML EP remove stale code for int64 via int32 double strides (#9959)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 111, "deletions": 391}}, {"node": {"oid": "1f5b0735085799ee8c93c65d4f9cc2b5cbc30cc8", "messageHeadline": "Minor DirectML EP provider factory comments (#9965)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "7d93498e0ec4b1b2a6f55161539513318a908e04", "messageHeadline": "[FIX] register softmaxgrad_13/logsoftmaxgrad_13 for rocm (#10177)", "author": {"name": "PeixuanZuo", "email": "94887879+PeixuanZuo@users.noreply.github.com", "user": {"login": "PeixuanZuo"}}, "additions": 194, "deletions": 14}}, {"node": {"oid": "6e88c11cae30c259ae4a7531d5943662c6fa2e32", "messageHeadline": "Refactor QDQ node group selection infrastructure (#10195)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 123, "deletions": 60}}, {"node": {"oid": "d52d3c00520a10075e3bc75fe90db5b4c83ee557", "messageHeadline": "Update C/C++ API docs automation to create a PR (instead of push to p\u2026", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 28, "deletions": 21}}, {"node": {"oid": "5ebb857501eb2b3145a42185f02bde7903862749", "messageHeadline": "Update onnxruntime_unittests.cmake (#10215)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "bacae967a21898141ac06d02cd9c9e6c30fbd139", "messageHeadline": "Update Cuda to 11.4.2, update architectures, support Ubuntu 20.04 (#1\u2026", "author": {"name": "vade", "email": "vade@vade.info", "user": {"login": "vade"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "2bbf1ac1e07586808ceae6e895678950acb05733", "messageHeadline": "Using better words. (#10210)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e7efcc93fe310dbc63f4d90ec33dbc9d9ec6303a", "messageHeadline": "[ROCm] update hipify-perl location (#10102)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 24, "deletions": 2}}, {"node": {"oid": "4ac327774334312a856716f8126fcc3fa5f0a521", "messageHeadline": "adding definition of concat operator for mapping it to onnx (#10062)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 37, "deletions": 5}}, {"node": {"oid": "cab4579b834c8fa53d23b961355209daecb14ad7", "messageHeadline": "remove six references (#9941)", "author": {"name": "Chris Hua", "email": "hua.christopher@gmail.com", "user": {"login": "stillmatic"}}, "additions": 9, "deletions": 12}}, {"node": {"oid": "0552a47ec2791a5f32fd768c6dad0cccb044117b", "messageHeadline": "Enable CUDA provider option configuration for C# (#10188)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 683, "deletions": 15}}, {"node": {"oid": "08f512b25e1624be166d6dd0284e2e737b8c994a", "messageHeadline": "Fix a Win GPU reduced ops pipeline (#10202)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "4ab891999a04e2aeee2377fb7fae69f2d6f6749f", "messageHeadline": "fix hardcoded type (#10205)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7b5464ed7bdd4eda5ea33977c008538b08da9025", "messageHeadline": "aten add_ op supports bf16 (#10084)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 112, "deletions": 26}}, {"node": {"oid": "34c025109c45f6dba9b642f5153686ade7d17b9a", "messageHeadline": "Exclude graph_runtime_optimization_test.cc from reduced ops build. (#\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "2803a9465db8a2f0d4e8fb4c783b8b687faf610f", "messageHeadline": "Add example of registering custom cuda op as shared lib (#10025)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 40, "deletions": 3}}, {"node": {"oid": "2078210a1cc2f9defc01452300ac582b0053f8cd", "messageHeadline": "Improve logging for symbolic shape inference", "author": {"name": "yz", "email": "zhanyao@microsoft.com", "user": {"login": "zhangyaobit"}}, "additions": 37, "deletions": 34}}, {"node": {"oid": "792db33f0140cf8c43a32d7cad7e8c8110a1fe49", "messageHeadline": "Enable loading of ORT format model graph runtime optimizations (#9901)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 844, "deletions": 518}}, {"node": {"oid": "97659495d982982a06eee6a4e3560923b988ff7e", "messageHeadline": "fix aten view op (#10050)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 50, "deletions": 20}}, {"node": {"oid": "91f85dfdada60e41a9d850e968fd9c58ae58facb", "messageHeadline": "update Dockerfile.manylinux2014_cuda11_4_tensorrt8_2 to TensorRT 8.2.\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c29397ad4f3549666b6fe35b1e8b972bec41d46d", "messageHeadline": "Modify the code to get correct ragne for symmetric quantization (#10170)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "0c517112c419627643fa4ba711f20cb0b783bdb5", "messageHeadline": "Automate Python API docs generation  (#10116)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 49, "deletions": 0}}, {"node": {"oid": "230f32360051bca8c376bd842e4387a5c29177ce", "messageHeadline": "add qdq support for LeakyRelu (#10077)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 98, "deletions": 36}}, {"node": {"oid": "1d3b34cc923a58fd138e32a110acb16aa564cbb9", "messageHeadline": "Add `.git` suffix to github URL.", "author": {"name": "Tongliang Liao", "email": "xkszltl@gmail.com", "user": {"login": "xkszltl"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "7208fcbe1c855c12d1f42aadb4093ab9ff3ead7e", "messageHeadline": "use wasmscalar as default kernel (#9988)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 12, "deletions": 9}}, {"node": {"oid": "28ce2a5a785804d6244d78274c6fbbda06f48b57", "messageHeadline": "Re-work hierarchy, fix virtual method overload/hiding (#10160)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 102, "deletions": 70}}, {"node": {"oid": "d5742f3a431bebc139318087ca126e1eb68fcaae", "messageHeadline": "moving from torch nightly build to stable build (#10150)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "3bc91c2151823830df83dbdc4a5d0a36df932b4e", "messageHeadline": "Move reduced ops files into build directory (#10030)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 174, "deletions": 184}}, {"node": {"oid": "a367f0664d831d4fb2557e8d63fc09d67d7386fe", "messageHeadline": "From Python 3.8 and on you need to explicitly add the current directo\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 15, "deletions": 9}}, {"node": {"oid": "3d6786c92e0abc672d23a8b17171c14c11521902", "messageHeadline": "update tensorrt multi gpu pipeline to tensorrt 8.2 (#10141)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "ceb17f82ffc9f72acb64f28629d7db93d98e5ba9", "messageHeadline": "Use FusedMatMul When Transpose is Between First Dim and Contiguous Ba\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 640, "deletions": 166}}, {"node": {"oid": "f780f06240ef407a838d7ea6d11a24cfcc436733", "messageHeadline": "ConcatGrad for OpSet13 (#10109)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 43, "deletions": 53}}, {"node": {"oid": "05d20343ee04990a7082ccb162faab9dd9a8305c", "messageHeadline": "Remove duplicated constant initializer copies for TensorRT nodes (#10\u2026", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 28, "deletions": 0}}, {"node": {"oid": "ce1a9ca6186c713b8962d1bf7cd6178cab21f64b", "messageHeadline": "Fix Microsoft.AI.MachineLearning NuGet App failure with multiple bina\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 752, "deletions": 54}}, {"node": {"oid": "7a1bdc2052bca1b4073229fee71123976731b866", "messageHeadline": "Don't check cache shape when using dynamic axis (#10090)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 16, "deletions": 15}}, {"node": {"oid": "4e9e01cb3c008335a2471c27dbdf7dd5d12e4224", "messageHeadline": "Fix SDL warnings in CPU EP (#9975)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1329, "deletions": 954}}, {"node": {"oid": "bd4fb4c5da0e774e5b7272a133904a31b7d60bfc", "messageHeadline": "Coding style fix. (#10080)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 5, "deletions": 7}}, {"node": {"oid": "cdbd678192b09655a14d63bc7f5f9005176cf5ea", "messageHeadline": "Check kMSDomain already exists before registering it (#10078)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 17, "deletions": 7}}, {"node": {"oid": "12ee2e942f4626d55f07725e4aaa0b8b346d5279", "messageHeadline": "add int8_t for Resize (#10067)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 132, "deletions": 5}}, {"node": {"oid": "4fd85cd97a82e2bf76e84425861deb99a42f9585", "messageHeadline": "Fix broken link to TRT doc in exception details (#9496)", "author": {"name": "Moshe David", "email": "mosdav165@gmail.com", "user": {"login": "mosdav"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d42feae042c3078cdf93b10865f78edf1a8c6a50", "messageHeadline": "Add citation file (#10061)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 15, "deletions": 0}}, {"node": {"oid": "f3c72de718f304102f1864c2dd5284c432f2e64a", "messageHeadline": "[QDQ] Add shared NodeUnit class (#10052)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 525, "deletions": 259}}, {"node": {"oid": "ef36488df024c96a37ab712a2a8de53176dbadcb", "messageHeadline": "Add BeamSearch operator for GPT-2 decoding (#9680)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3262, "deletions": 3}}, {"node": {"oid": "fab39b470439423ee2ee7fcad70bd122055183c7", "messageHeadline": "Update optimization level message in perf_test tool (#9972)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "102f9b05e17366e7fa474fb9faed6371c09ac644", "messageHeadline": "Support new symbolic function api from PyTorch with PythonOp (#9880)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 12, "deletions": 1}}, {"node": {"oid": "93636cbd207ab9f28ebdc6deecaa320f05240d95", "messageHeadline": "Reduce ops for DNNL ep (#10056)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 448, "deletions": 171}}, {"node": {"oid": "44c701192b3ea3f927f9eaee7a05b30b87ba9b1f", "messageHeadline": "Revert a bad change in bfc_arena.cc (#10057)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6357c1297798705f173a0b990e7a118233b8cfa0", "messageHeadline": "use inplace reshape (#9991)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 21, "deletions": 9}}, {"node": {"oid": "7922a8c22f42d06abff2682b047e62f604e9755e", "messageHeadline": "Optimization Convolution op when using dnnl ep (#10051)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "3466ee45a3e851e176431c49a0a6ef35e35efaef", "messageHeadline": "Add hash value typedef. (#9710)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 70, "deletions": 52}}, {"node": {"oid": "4e73cc83d6d646f2c8ee52ba3ee90dc5669b1bef", "messageHeadline": "Fix building DNNL EP with clang (#10014)", "author": {"name": "Chih-Hsuan Yen", "email": "yan12125@gmail.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b327e89efa22c5c8c2c32988dbbe60154df2a39e", "messageHeadline": "Standalone TVM Executor Provider (#10019)", "author": {"name": "Valery Chernov", "email": "black.chervi@gmail.com", "user": {"login": "vvchernov"}}, "additions": 2251, "deletions": 7}}, {"node": {"oid": "16274beb6fb01b054c7729859c63e7c93574b747", "messageHeadline": "update TensorRT EP to use TensorRT 8.2 (#9981)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 271, "deletions": 238}}, {"node": {"oid": "ee975de77b547554c99788e30a1779ebc03d6c8c", "messageHeadline": "reorganize quantization files (#10023)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 13, "deletions": 12}}, {"node": {"oid": "6cdab0625523d3c800ee60e583ed01818fc989dc", "messageHeadline": "Enable argument files in build.py. (#10040)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 11, "deletions": 3}}, {"node": {"oid": "20f8a06f1f111ab4915c55c53338e67122f8c5b8", "messageHeadline": "Remove OpenMP code (#10032)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 287, "deletions": 330}}, {"node": {"oid": "8043a9facc311ad11d39f2e614103c82eeb25809", "messageHeadline": "Bump master version to 1.11 (#9957)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 8800, "deletions": 8792}}, {"node": {"oid": "91096781c34a5278c496bbd94fd0da4445a3e005", "messageHeadline": "A small fix to allocators (#10042)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9d9ebd3b8536cc93fe8ac32acdffcd490c16c70c", "messageHeadline": "Fix some static analysis warnings in the core framework (#10033)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 201, "deletions": 140}}, {"node": {"oid": "e0a0f385bb10b5a0c98a8dfe52921de64f0bee51", "messageHeadline": "Fix some warnings in mlas (#10034)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 66, "deletions": 27}}, {"node": {"oid": "af71da0ac68ad036ae3748c53f47c06bf6ea69b7", "messageHeadline": "Yield op supports bf16 (#10035)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "703becd796ca7359b6424f74e1884e6de1780dd4", "messageHeadline": "Fix a bug in fusion_embedlayer.py (#10022)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5be0fa13c0d3fd60ace846c5c2394b16ff6c2e01", "messageHeadline": "[DML] Fixed huge bug in ORT_NO_EXCEPTIONS for DML back end, the check\u2026", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9e04b7e59b3ee1d0eff70010077f659a4e32a86d", "messageHeadline": "Remove memcpy in in-place ATen ops (#9913)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "a7c2d1cb092db83e427a0e9a27282ae34d8a7029", "messageHeadline": "bf16 for dlpack (#10016)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "cd0af7ad447a116b6084d0dac2044fb91d8d0f5d", "messageHeadline": "Symmetric quantized convolution kernel ARM64 (#9772)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 2258, "deletions": 28}}, {"node": {"oid": "7e55a942cd6d06421d9a1f90d52dbfaf19bd4da1", "messageHeadline": "Add torch 1.10 requirements for rocm (#10028)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 12, "deletions": 0}}, {"node": {"oid": "6de2a878cbdda075662ee2ea92444c59814f9d30", "messageHeadline": "[js/react_native] Fix a broken manual build (#10012)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 31, "deletions": 19}}, {"node": {"oid": "7b63d1102b0fbbe79cdd55ad4daea0ffaa841ff1", "messageHeadline": "Fix some warnings in orttraining code (#10009)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 249, "deletions": 240}}, {"node": {"oid": "c82160bbd0e74a446dab17440dec87cdaaab5d6f", "messageHeadline": "Add AtenOp at:bitwise_or (#9662)", "author": {"name": "Gani Nazirov", "email": "ganinz@hotmail.com", "user": {"login": "ganik"}}, "additions": 47, "deletions": 3}}, {"node": {"oid": "ad99dff2980285ff064d9c846b0ac3d47f14bbfb", "messageHeadline": "POWER10: Update builtins for DGEMM", "author": {"name": "Rajalakshmi Srinivasaraghavan", "email": "rajis@linux.ibm.com", "user": null}, "additions": 90, "deletions": 83}}, {"node": {"oid": "5d821b5bd9d1db0660975c4bd2416975a629f852", "messageHeadline": "Address null dereference warning in div_grad_impl.cu. (#10010)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 41, "deletions": 41}}, {"node": {"oid": "777a80fbc15eaf1b79e06b86dcd57cdf1d156f50", "messageHeadline": "Abjindal/eager onnx operators fix (#9968)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 56, "deletions": 29}}, {"node": {"oid": "d0b08af37a07cc6ad270e4af0a89d6047ad1fe4c", "messageHeadline": "Implementation of  QAttention for the DNNL execution provider (#10004)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 975, "deletions": 32}}, {"node": {"oid": "787755328b98d8830c421f3c7865767c9a2ab570", "messageHeadline": "Add s8s8 for depthwise qconv 3x3 5x5  (#10008)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 991, "deletions": 20}}, {"node": {"oid": "b4434c76946eaabb5310b1444309c49e401bd9cb", "messageHeadline": "Automate generation of C/C++ API docs (#9997)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 32, "deletions": 2}}, {"node": {"oid": "b000ec91cc3141a43ba7a136a8ce1b2bc82e6341", "messageHeadline": "Add quantization tool and its unittest with s8s8 support (#10007)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 415, "deletions": 173}}, {"node": {"oid": "7a70d22150b383b342a849143549d10e69aeab54", "messageHeadline": "Change some const to constexpr in unit tests code(#10002)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 390, "deletions": 374}}, {"node": {"oid": "ffdafb2012d6a651c8f440ac486f9d0d5a939901", "messageHeadline": "add fallback of s8s8 support on x64 (#9995)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 95, "deletions": 312}}, {"node": {"oid": "3c79f3055f4b9d3a67c7ee814f8ebc9cba1bda12", "messageHeadline": "Weaken x86.get_pc_thunk for NDK <= r22 (#9994)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "343a76945be1323e14c5251eb5302037c7d77eca", "messageHeadline": "Fix some documentation errors plus ones generating doxygen warnings (\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 22, "deletions": 20}}, {"node": {"oid": "4669048b478172914fa9611df5b5d042bc208f45", "messageHeadline": "Handle compiler warnings for TRT EP (#9956)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 93, "deletions": 74}}, {"node": {"oid": "e0960d7d79ad25d7f50036ea81342265affd1e04", "messageHeadline": "Change assert on a null value to an ort_enforce (#9982)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b9909f985e6794ba091c490b683c852611eeb2a1", "messageHeadline": "[js/web] rename build-def.ts to build-def.d.ts (#9954)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "c7e8365e1fff0ebdd029b81477422dbc2afa67e2", "messageHeadline": "Resolve security issue reported by dependabot (#9983)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 869, "deletions": 795}}, {"node": {"oid": "2fdcaffbb726398be80e8f21f37636c1a0458e54", "messageHeadline": "add missing test files (#9989)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 91, "deletions": 0}}, {"node": {"oid": "41fd745996d2b7f949d67b985fad8f57a3b7ff07", "messageHeadline": "Fix build error when using '--build_minimal extended' and '--build_wh\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "a7abd541c79748d6fc6ba596ddb21ccdcb3c1ce1", "messageHeadline": "Correct message type (#9973)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "42c176b60c502ab3a26f7289397fa5cad5becba7", "messageHeadline": "Update default opset to 14 in ORTModule (#9743)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 417, "deletions": 93}}, {"node": {"oid": "bc0f2d173ace7be561c4b828f93fa55574205302", "messageHeadline": "enable s8s8 QDQ fusion on ARM (#9961)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 550, "deletions": 201}}, {"node": {"oid": "cd552e1bdad6bb57da0acb13d76dc7ecb4d0be20", "messageHeadline": "Add build.py option for disabling memleak checker", "author": {"name": "Martin Ho\u0159e\u0148ovsk\u00fd", "email": "martin.horenovsky@gmail.com", "user": {"login": "horenmar"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "051d0059261c2fca619faf975aca238ef95c6670", "messageHeadline": "Fix build warnings with VS 2022 (#9967)", "author": {"name": "Nick Kreeger", "email": "nick.kreeger@gmail.com", "user": {"login": "nkreeger"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "0c72f1cd5a51fbe79b3755a993912d362eef35f5", "messageHeadline": "add copyright (#9943) (#9970)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 89, "deletions": 0}}, {"node": {"oid": "35cf8b8725c41bae7d82f1331fb97cda122a80d7", "messageHeadline": "[js/react_native] npm audit fix (#9876)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 1191, "deletions": 1953}}, {"node": {"oid": "0adeb86bfda43b42b3351519a648a1824cbdbfff", "messageHeadline": "Fix ortmodule for the pytorch model with ort device (#9927)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 119, "deletions": 36}}, {"node": {"oid": "fb30e9fdae53c6ca0d4068a5f0656205e953611d", "messageHeadline": "Remove /safeseh link option from non-msvc builds (#9744) (#9935)", "author": {"name": "Patrik Vavercak", "email": "56153264+pvavercak@users.noreply.github.com", "user": {"login": "pvavercak"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9e7d52a80169847071ee35a135e9ff97e7dc38fb", "messageHeadline": "Update parity_check_helper.py (#9884)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f60a287a64d509ae60c2ce1190effe9e8bdb0f4e", "messageHeadline": "Add __x86.get_pc_thunk.bx to avoid dependency (#9955)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 44, "deletions": 0}}, {"node": {"oid": "a7f649db7c28d5d2b532b2b9a2d1a191a90b67f8", "messageHeadline": "Enable proper override using MIMalloc (#9944)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 167, "deletions": 282}}, {"node": {"oid": "94dc8440394bbb7702c2b43f1609eb9a12aa0840", "messageHeadline": "Merged PR 6743839: Merge latest ORT master", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 7631, "deletions": 32693}}, {"node": {"oid": "b34b991aea41fc2fdc758fc363db1f167fa3b763", "messageHeadline": "Improve reduced ops and types build (#9908)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 243, "deletions": 130}}, {"node": {"oid": "28cc3b63a05d4c6169848efe3451bb6761c6034a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/justoeck/mer\u2026", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 7631, "deletions": 32693}}, {"node": {"oid": "840212e115f7468fc0968570e9c0b5c860fd1fce", "messageHeadline": "Enable OneHot kernel for ROCm EP and add Dockerfile for ROCm 4.3.1 (#\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 177, "deletions": 5}}, {"node": {"oid": "63c888994490d3800c73b38ea70a27da1916cd98", "messageHeadline": "Restore arm64x onnxruntime binaries (#9950)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 26, "deletions": 0}}, {"node": {"oid": "5e4d58a50abb261582c438dadafde8cf8c41f265", "messageHeadline": "Openvino ep nuget (#9909)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 32, "deletions": 3}}, {"node": {"oid": "a23cd5b697a2500035420ea99fd0d98f35d421cb", "messageHeadline": "Update Xamarin sample code (#9925)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 98, "deletions": 31470}}, {"node": {"oid": "58728f95a297fcfa24783f85ecccc54fb8248ac7", "messageHeadline": "Fix conv quant reduce range option (#9922)", "author": {"name": "MengX1991", "email": "93172434+MengX1991@users.noreply.github.com", "user": {"login": "MengX1991"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "5871ca1cd16ed8e91411a4de51425bb312be789c", "messageHeadline": "Comment out unused parameter (#9914)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "67a30ef71655665a3b2b48d4650aa70b55c50edd", "messageHeadline": "Address some code scan issues (#9873)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e61301917442486e6ef9c7f39cddab72739f0fc5", "messageHeadline": "add s8s8 support for quantized conv and gemm (#9902)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 4713, "deletions": 652}}, {"node": {"oid": "d8c71304c1288d343858d0e48f7cda2b071c7947", "messageHeadline": "Update Dev_Guide.md (#9921)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "02aa16e3ea2faa2a893d279c8e323ec3f672bc0b", "messageHeadline": "QDQ tool modification part3 (#9904)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 34, "deletions": 16}}, {"node": {"oid": "4ff78aae45c102a4be81594d4fffd3bae3115af7", "messageHeadline": "Merge pull request #9917 from microsoft/user/dwayner/FnsCandyToleranc\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 65, "deletions": 39}}, {"node": {"oid": "5edaa75ef65af93a06c74bcc602d446b8fd63478", "messageHeadline": "Fix LoadFromStream to not use wss::Buffer internally (#9918)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 514, "deletions": 77}}, {"node": {"oid": "8db49e3d0f9bb60957603d1bcbff7032e37a9086", "messageHeadline": "add ortmodule and eager mode test (#9888)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 105, "deletions": 5}}, {"node": {"oid": "6e4c534ce219f015024ecb60ecc7b08cba4b3981", "messageHeadline": "Relax tolerance slightly more for Intel after autopilot run", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1c38ceda49394564cee6b168601d3bb612228d3e", "messageHeadline": "Add fusion support for Dnnl execution provider (#9897)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 1597, "deletions": 98}}, {"node": {"oid": "06e63218be341c4297ae81e9473d8531d790911b", "messageHeadline": "changing commit for windows build for eager mode (#9912)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "77e67a6de74fe05485e983b574f4b91540c8e443", "messageHeadline": "Add one more example line", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "ef7671b938e10d2b4a9b39d368d13cdeea81e4b1", "messageHeadline": "Comment out old lines", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "912e50f61c7faf8a25c6d262717ce4d0bc120ee8", "messageHeadline": "Add CI minimal build with all options disabled. Fix python binding co\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 438, "deletions": 324}}, {"node": {"oid": "5190953f6b665148de03c63118255c142f0d38ab", "messageHeadline": "Merged PR 6728025: Merge latest ORT master branch", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 529, "deletions": 167}}, {"node": {"oid": "2e9dc8a0e2e76dac3bf56baf57aaad1e61335c40", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into tmp_merge", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 529, "deletions": 167}}, {"node": {"oid": "5ba40795466058fc68ccae8180f40fa6567e10c9", "messageHeadline": "Merge branch 'DmlDev' into user/dwayner/FnsCandyTolerance30696168", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "7a3abd863fa6b99f966ac4134e5b6ae6f9f20527", "messageHeadline": "Update WinML model test tolerances for tiny_yolov2 and FNS_Candy", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 65, "deletions": 39}}, {"node": {"oid": "3f5c1e1c58939c0de30e232ea0ef0c30644166b0", "messageHeadline": "Update to include the Xamarin targets for internal ORT builds so the \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "8d88a6ac7f15cfb0f075fb420ec24680872cd60d", "messageHeadline": "add --amdgpu-target=gfx90a (#9820)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "00c979db4db061ab7d9836d6a3d8f797087a3f15", "messageHeadline": "Update doc for operators/opsets supported by mobile package (#9899)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 97, "deletions": 92}}, {"node": {"oid": "078782ea3c27b3d1e26bf621fe2ffb3ca012d439", "messageHeadline": "exclude test case from reduced-op build (#9895)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "3f82b2d5b13848c29b8f0d8bc679d20dea20b265", "messageHeadline": "Merged PR 6723735: Need to unsquash the last RI to maintain history", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "73d106e33a666a49c42ac8c0925e19680fa1386f", "messageHeadline": "Merge commit '9345894c823de67f913d907da91dbc07ac322041' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "175acf08f470db0bb2e4b8eefe55cdeb87c8b132", "messageHeadline": "ScatterND supports negative indices (#9739)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 82, "deletions": 12}}, {"node": {"oid": "c161813217d9bb17c52611e284f6956a7b30de37", "messageHeadline": "Misc InstanceNorm CUDA kernel changes (#9879)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 321, "deletions": 39}}, {"node": {"oid": "6de79d82c85f1acff58d0ba31ef48787b0e02152", "messageHeadline": "Fix Training Packaging pipeline (#9885)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 18, "deletions": 18}}, {"node": {"oid": "d8a7e1d159e22389d640ced85c6cd541acc812ce", "messageHeadline": "Merged PR 6718335: RI 11/30 from github", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 9779, "deletions": 4225}}, {"node": {"oid": "740679d3290b0df594d2e14be959dfbdb405f80b", "messageHeadline": "Abjindal/fix windows ci pipeline (#9883)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "9345894c823de67f913d907da91dbc07ac322041", "messageHeadline": "Add build option to enable cuda profiling (#9875)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 24, "deletions": 7}}, {"node": {"oid": "16bfd3c771c417fa047aa0e551dc493818c2bb4c", "messageHeadline": "Cancel transpose optimizer for resize (#9870)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 276, "deletions": 232}}, {"node": {"oid": "0baf687f2df01bab8f3160c464b26ac383d9dff3", "messageHeadline": "QDQ tool modification part2 (#9720)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 246, "deletions": 23}}, {"node": {"oid": "57a6f7c2051b5f161d950bb421432efa30d1422c", "messageHeadline": "Various fixes to fix WindowsAI RI build. (#9877)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 9, "deletions": 8}}, {"node": {"oid": "a0afd7303da0c6622c1d73bd6dda1bec1fae457c", "messageHeadline": "add int8_t support for pool operators (#9852)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1224, "deletions": 417}}, {"node": {"oid": "ec9b0ed800650113a35be5ecf2d68394dbfe6ad1", "messageHeadline": "[python manylinux package] emit warning if missing CUDA/TensorRT depe\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "27e337ed7a0ca55126ac2f3a250a5ab16e54ea63", "messageHeadline": "[js/node] npm audit fix (#9861)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 140, "deletions": 459}}, {"node": {"oid": "bf716e667c43d5091af5bd2418eb7bdbcb9d4c23", "messageHeadline": "layernorm throw error if input has no data (#9837)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "9e75ebf0dc77b65eb7729f14638a5df62d6c2e7d", "messageHeadline": "Remove redundant inline specifiers, sync server IsLittleEndianOrder w\u2026", "author": {"name": "cristei", "email": "51248378+cristeigabriel@users.noreply.github.com", "user": {"login": "cristeigabriel"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "37bf46eb19b2b18b7dccd861af6ab0fdd585eda4", "messageHeadline": "support print in ort eager mode (#9825)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 1137, "deletions": 489}}, {"node": {"oid": "1e9e57df3efddc2fc41540e79149e670ac5304c6", "messageHeadline": "no fallback when enforcing explicit EP registration. (#9863)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "a3ebc5e08290b4000e8509d21819d7f486e23a14", "messageHeadline": "[js/web] do not use nodejs type 'Buffer' in web (#9839)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 20, "deletions": 17}}, {"node": {"oid": "6eb0c8d4202d265951f5c165f0d82fff3bf81413", "messageHeadline": "fix build break in release pipeline for Node.js binding test (#9850)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "53c43e99498d792ecd3bea158e95a51e4db94ae0", "messageHeadline": "WinML RT API: Add PixelRange Metadata to Bind() call PropertySet (#9827)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 38, "deletions": 4}}, {"node": {"oid": "18fd2cf4571061b61115b47b524ae6828057f95e", "messageHeadline": "Fix potential data race with OrtValue usage in Python (#9841)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 52, "deletions": 3}}, {"node": {"oid": "0ae0f29f140c5d7b4077df024da75abd30367e58", "messageHeadline": "[OpenVINO-EP] V3.4 Release with OpenVINO 2021.4.2 LTS Release (#9848)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 26, "deletions": 13}}, {"node": {"oid": "926109bd638c3d27dfa6077344bdaec8f6814f01", "messageHeadline": "Added algorithm for std::clamp (#9797)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "d012d9fb913230205620771c94cb6f5cb89a4203", "messageHeadline": "Fix memset size (#9840)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6749e9fd4456c895df02b456f20be95014d84dfc", "messageHeadline": "Cuda instance_norm fix (#9826)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "24f3d72b77a475d915c16a86aa8f11367d4bba21", "messageHeadline": "relax atol and rtol for einsum ut (#9842)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "8564fc193317ab31e209a67a9aaeb34eca991699", "messageHeadline": "POWER10: Add optimized dgemm kernel (#9652)", "author": {"name": "RajalakshmiSR", "email": "58744256+RajalakshmiSR@users.noreply.github.com", "user": {"login": "RajalakshmiSR"}}, "additions": 799, "deletions": 599}}, {"node": {"oid": "bf5e9a5044c3f5b2298080646414fcd5034f8ffe", "messageHeadline": "bumping up ORT_API_VERSION to 10 (#9838)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fb4a8e12fc3b0eb193a6fe88e6e99705b390955c", "messageHeadline": "Limit inclusion of Xamarin mobile target frameworks. (#9834)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 317, "deletions": 37}}, {"node": {"oid": "74ca417c0e0e8ce4fd78d727e323f77194cab26b", "messageHeadline": "[js/web] optimize bundle file size (#9817)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 382, "deletions": 70}}, {"node": {"oid": "bcc6ab29f6307ebac3b72d556f78bc86d731846d", "messageHeadline": "Trim DataTypeImpl binary size (#9813)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 118, "deletions": 174}}, {"node": {"oid": "567749b2dc57bce6412d05b3f40a88425dbbf9c9", "messageHeadline": "Expose IOBinding SynchronizeInputs/Outputs via C/C++/C# And Python AP\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 212, "deletions": 34}}, {"node": {"oid": "8dcd388ec5055b0cb7a92654d9933c16a8b8fa6b", "messageHeadline": "Bump ansi-regex from 5.0.0 to 5.0.1 in /js", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "c779f2cad5cec221a539c44a497894270a804b61", "messageHeadline": "Bump path-parse from 1.0.6 to 1.0.7 in /js", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "88d8d2d02aafdb49aba88f8a1f8ea39df1660337", "messageHeadline": "Remove Copy NuGet Package to Azure Blob Store build step (#8991)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 24}}, {"node": {"oid": "efcdbac4e76fcd865117d9437d86ea7f7751d767", "messageHeadline": "Fix ONNX Runtime Mobile link in js/web/README.md (#9828)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3a85ade5116830792b9ad2497f309505e58a5e14", "messageHeadline": "Avoid segmentation fault when creating session with TRT EP or OpenVIN\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 21, "deletions": 10}}, {"node": {"oid": "7396689c2b358b4b73aa0fd4ecbd4253d4e6d770", "messageHeadline": "Merge pull request #9765 from microsoft/user/dwayner/DML1.8forORT1.10", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 2300, "deletions": 1083}}, {"node": {"oid": "6856619b18d61632a3f07fc02b2ceaaf37ecd0a2", "messageHeadline": "Decoder Attention CUDA Op (#9792)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2045, "deletions": 259}}, {"node": {"oid": "16ddaf564c14436a3978a3f3f3316b899977a5cd", "messageHeadline": "Fix uninitialized warning by some compiler. (#9822)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "f28d7eca67043ef620c907b3d70114f475b7cd84", "messageHeadline": "Disable DML dynamicquantizelinear in backend tests DML Bug 33073263", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "1b953c64237d619ee58422bf1d47cef89c05dd35", "messageHeadline": "Fix some code defects (#9810)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "d1b772e09afd41f03845dfb6ea9d13cea0c297a6", "messageHeadline": "Mark dynamicquantizelinear as broken in DML", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "ba339e667bb9f7832797b8775de66ef8a7bda4ce", "messageHeadline": "Add training performance investigation script (ONNX graph analyzer) (\u2026", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 85, "deletions": 0}}, {"node": {"oid": "f047be55d4115c5ac208bf5c9f4aae0744834b20", "messageHeadline": "Merge branch 'master' into user/dwayner/DML1.8forORT1.10", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 274, "deletions": 66}}, {"node": {"oid": "7e026286aec717a46597f0d7bc261262a1eb6517", "messageHeadline": "Safe remove quantized inititalier which may still be used (#9788)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 137, "deletions": 42}}, {"node": {"oid": "e520bb5145776b2f4a23972ad0f6741c4c1b108b", "messageHeadline": "Improve print functions for NodeArg, Node, and Graph (#9801)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 84, "deletions": 21}}, {"node": {"oid": "9d3c63263b74cdb68e13f748fe63d0acd04ef7a3", "messageHeadline": "symbolic_shape_infer: Improve error message on mismatched types (#9809)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 53, "deletions": 3}}, {"node": {"oid": "289b1bdc86d2cc2ff2ee4921804c94a3491b3b8d", "messageHeadline": "Merge remote-tracking branch 'ado_wai_ort/DmlDev' into user/dwayner/D\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 2295, "deletions": 1080}}, {"node": {"oid": "32419974adb6794b31b30b0005143e7456641bf9", "messageHeadline": "Merge remote-tracking branch 'origin/master' into user/dwayner/DML1.8\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 6081, "deletions": 2721}}, {"node": {"oid": "e0ffc30a0bfd68199c07250fe204e5a7e197ed17", "messageHeadline": "Update to 1.8.0", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "afd60a274cfa1af3cd00f66f12afeeb26febb629", "messageHeadline": "Fix some places where there were unused parameters when sparse tensor\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 88, "deletions": 8}}, {"node": {"oid": "b77a6cda5012120d09e92e80086357163dbde513", "messageHeadline": "squeeze and unsqueeze operators for the dnnl execution provider (#9798)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 371, "deletions": 17}}, {"node": {"oid": "df5eb98b6e21aadcf57be1e0a7227267b61dbd7a", "messageHeadline": "Bump tmpl from 1.0.4 to 1.0.5 in /js/react_native", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "fcc167dd47371b0cfafdc4dcde6eb13f652d8f67", "messageHeadline": "fix reshape implementation in eager mode (#9741)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 19, "deletions": 5}}, {"node": {"oid": "7ea19539f83aaac5d394d7fd573dd4c774cf031d", "messageHeadline": "Bump path-parse from 1.0.6 to 1.0.7 in /js/react_native", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "0556b41127d0ebf9fd400ae5ffe74405a4f9559c", "messageHeadline": "Bump react-native from 0.63.4 to 0.64.1 in /js/react_native", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 806, "deletions": 872}}, {"node": {"oid": "6977804804f0c52979b5267fa6cee1f7779ffebd", "messageHeadline": "Bump electron from 12.0.2 to 12.1.0 in /js/web", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 21, "deletions": 28}}, {"node": {"oid": "3af14fc55404f10380bd651554656f3d4bd1a1b1", "messageHeadline": "Updated SoftmaxGrad and LogSoftmaxGrad to support version 13. (#9733)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 268, "deletions": 84}}, {"node": {"oid": "dc1724b0e272b2804fa8eb7ddd5ec700efdc7150", "messageHeadline": "Reduce DataTypeImpl binary size (#9783)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 101, "deletions": 125}}, {"node": {"oid": "f390347c11c3aa4a5c0481b5cbf7cf1ec45b7da3", "messageHeadline": "Add CUDA Kernels of RandomNormal[Like], RandomUniform[Like] (#9761)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 602, "deletions": 4}}, {"node": {"oid": "6284cbe833bc47eb90ed4082ff2ae015b8e5c004", "messageHeadline": "Add TensorShape noexcept for move ops and fix some warnings (#9802)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "a2787ca67ca770f12493106bac493ae8f3d657ad", "messageHeadline": "Merged PR 6688023: ORT GitHub to DmlDev RI", "author": {"name": "Numfor Mbiziwo-Tiapo", "email": "numform@microsoft.com", "user": {"login": "nums11"}}, "additions": 9427, "deletions": 2779}}, {"node": {"oid": "533b20c6ca7b36321ed3fdae37f3dbe1876c8baf", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into dmldev_temp", "author": {"name": "nums11", "email": "numsmt2@gmail.com", "user": {"login": "nums11"}}, "additions": 9427, "deletions": 2779}}, {"node": {"oid": "8ef6aff7343aff9bf1533de34ca890ff16be9de3", "messageHeadline": "Zhalei/dwqconv3x3 5x5 arm64 (#9714)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1119, "deletions": 20}}, {"node": {"oid": "7242627fecb5074aa6c0a84aae2a054ba53519ea", "messageHeadline": "Integrate TensorRT into GPU Python package (#9785)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 209, "deletions": 167}}, {"node": {"oid": "76715ad52560d762f1cfafb2202c962e17d85c44", "messageHeadline": "Delete ioscross code (#9793)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 32, "deletions": 90}}, {"node": {"oid": "1aa21df149fe7be5ac39b3d23f80234a6f4d7890", "messageHeadline": "Fix issue with debug VS2022 build when python bindings are enabled  (\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 30, "deletions": 14}}, {"node": {"oid": "191bc518a16bbc52dfccfe444149fa2674dc4f69", "messageHeadline": "Merged PR 6676705: Add opset 13 for ops which are unchanged", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 795, "deletions": 630}}, {"node": {"oid": "e23892ddbecbc2be5a527c66c105f4949c742236", "messageHeadline": "Support disabling support for the optional type in ORT builds (#9745)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1558, "deletions": 1223}}, {"node": {"oid": "9fb3fac5a0c7bb95d5c08ba11ff8cc7fcb671c2a", "messageHeadline": "Enable LearningModel::LoadFromFilePath in UWP Apps (#9790)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3654a5d60e345d662c8a9566cd8e922efd82dbb0", "messageHeadline": "Register Custom Symbolic of torch.einsum for ORTModule (#9590)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 522, "deletions": 1}}, {"node": {"oid": "6545e24b6059afb3e0883d29f2203c1274e30a28", "messageHeadline": "Update mobile prebuilt package ops to add support for opset 14 and 15\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 251, "deletions": 16}}, {"node": {"oid": "20ec48129ba98d0697d5e057b358fa406020ae61", "messageHeadline": "fix typo in LearningModelDevice.cpp (#9787)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "871d3fb839a846a94ae1f7a20b0c97d5496c80f1", "messageHeadline": "Fix a bug in ReluClip fusion (#9764)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 33, "deletions": 1}}, {"node": {"oid": "b409cbe62c2f33a0eea19e027876d9856aa1515d", "messageHeadline": "Fix incorrect library reference in Python manylinux package for CUDA \u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "421e4c03cec43a53a800e18d48a4c9b6edbccfee", "messageHeadline": "Update default cast propagation strategy from None to FloodFill (#9713)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 26, "deletions": 21}}, {"node": {"oid": "9acbfeba0962370ea7299770c28e2d5f009b9dea", "messageHeadline": "Address some code scan issues. (#9752)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "99afb87a020602bdea59cb21895b69a80facb474", "messageHeadline": "Update DirectML 1.5.1 to 1.8.0 for ORT1.10", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "ac57afc3a69dfc6a57caeee678966a1fce50bca9", "messageHeadline": "Update ONNX to 1.10 globally in CIs (#9751)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 12, "deletions": 7}}, {"node": {"oid": "1d03baa8cc3acd52e1d84b0073f1daa5e11cfcbd", "messageHeadline": "Openvino ep 2021.4 v3.3 (#9588)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 266, "deletions": 61}}, {"node": {"oid": "4ca11b05a540a29351048224584894e31b829ba4", "messageHeadline": "Remove python 3.10 from rocm docker image (#9749)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 14}}, {"node": {"oid": "1c84621020f46ad5c27fd793c2a7df458b0e2c37", "messageHeadline": "Adding ARM64 depthwise convolution kernel for symmetric quantization \u2026", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1530, "deletions": 15}}, {"node": {"oid": "9f4e8cf6a0c27e3b3723b2df8b22c827bfdb06ec", "messageHeadline": "Update training pipelines to pytorch 1.10 (#9709)", "author": {"name": "raviskolli", "email": "48601275+raviskolli@users.noreply.github.com", "user": {"login": "raviskolli"}}, "additions": 231, "deletions": 27}}, {"node": {"oid": "99257eb8e3d528f05fa4248df2c843147d795920", "messageHeadline": "support build option to include external graph transformers  (#9478)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 330, "deletions": 1}}, {"node": {"oid": "6e09fc5152a109e46f22063ac9f8a1b82b482fda", "messageHeadline": "Implement block wise softmax for reduction dimention > 1024 cases. (#\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 841, "deletions": 342}}, {"node": {"oid": "3d0bd2596fe876b665eb9f5a0ca688dbebccbe31", "messageHeadline": "Enable creating OrtValues from ID3D12Resources from the onnxruntime C\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 295, "deletions": 203}}, {"node": {"oid": "21eb747a0fcad6559c4e601697e2f4fa06e2fb27", "messageHeadline": "Custom thread creation and join hooks (#9426)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 386, "deletions": 43}}, {"node": {"oid": "52203e06dd422ae2b705c4465af2e42415f4f3cf", "messageHeadline": "Merged PR 6668707: Fix Resize float16 ROI tensor (part 2)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 98, "deletions": 98}}, {"node": {"oid": "9f69d8bbae3c39c9e83add416d3e55d607357b90", "messageHeadline": "Disable partial runtime optimization implementation by default (#9748)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 19, "deletions": 23}}, {"node": {"oid": "a17bdaf72512260657c2bb3785c44782478f83e7", "messageHeadline": "Enable JoinModels API in WinML+RT Experimental API (#9746)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 507, "deletions": 19}}, {"node": {"oid": "3e5d636f6ce355e6afd7946ee0634dfe382d8866", "messageHeadline": "Merged PR 6668481: Fix Resize float16 ROI tensor", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 47, "deletions": 4}}, {"node": {"oid": "5ad6dbb314f3a2dd10450273ceeeb0513cad79b3", "messageHeadline": "Remove experimental from ORT format namespace (#9729)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 389, "deletions": 422}}, {"node": {"oid": "f6edf135132bbcbeb8e2294078d305c4353422cc", "messageHeadline": "Implement a Gemm/Sum fusion pattern (#9699)", "author": {"name": "Aidan Beggs", "email": "nadiasggeb001@gmail.com", "user": {"login": "quantatic"}}, "additions": 749, "deletions": 12}}, {"node": {"oid": "997266a62020d6ba1d303928a16bf53de09c76b0", "messageHeadline": "Add build.py option to disable ORT format model runtime optimization \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 94, "deletions": 39}}, {"node": {"oid": "363cb2300b31a7614ff42e1e3d5ec3163ae7a098", "messageHeadline": "Merged PR 6664646: Merge ORT Github and address DMLXP build error fro\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 3535, "deletions": 1833}}, {"node": {"oid": "2a9208188f0496f40e696f65f0541498b1f1d4cb", "messageHeadline": "ORT TensorShape changes to avoid build error in tools overriding GSL \u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "b9c4b2e0007271cac9d41aba675bd627ba0d7931", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/jeffbloo/Mer\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 3530, "deletions": 1833}}, {"node": {"oid": "93e239747f3073cb3a1ff41381feea7433fcbbee", "messageHeadline": "Construct valid graphs for ONNX checker for IR version < 4. (#9665)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 79, "deletions": 1}}, {"node": {"oid": "32c896df6d29822a027663a30582cbbacf794b81", "messageHeadline": "Add DynamicQuantizeLinear to DNNL EP (#9620)", "author": {"name": "zhaoyang-intel", "email": "zhaoyang.wang@intel.com", "user": {"login": "zhaoyang-intel"}}, "additions": 291, "deletions": 4}}, {"node": {"oid": "559adbb534631b73ad68ae42c5885154167c4057", "messageHeadline": "Replace deprecated PyEval_CallObject() call with PyObject_CallObject(\u2026", "author": {"name": "Nick Kreeger", "email": "nick.kreeger@gmail.com", "user": {"login": "nkreeger"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "724009289b74315ea70d856c19e9353a080b9086", "messageHeadline": "Fix Issue #9671  (#9691)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 29, "deletions": 19}}, {"node": {"oid": "03f9d77e173289b80cf53178480d192206a7ec4a", "messageHeadline": "Fix index out of bound bug of generated Gather (#9673)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 42, "deletions": 3}}, {"node": {"oid": "e6f0fdd6539eab7d42b76a826ca9090272d4f0ef", "messageHeadline": "Strip AMD libraries bundled with Python package due to libonnxruntime\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 3, "deletions": 1}}, {"node": {"oid": "acb54592683adb52bed3c16a75718a46b2b84b16", "messageHeadline": "Random CPU kernels should respect global seed (#9726)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 20, "deletions": 29}}, {"node": {"oid": "9d84811fb67ad5f434abaf707146ba1019a7b953", "messageHeadline": "fixing pypi pipeline for release (#9716)", "author": {"name": "Viswanath Boga", "email": "44417868+viboga@users.noreply.github.com", "user": {"login": "viboga"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "962feff7f1442aa5a4806a92020fce609619f991", "messageHeadline": "Update linux-ci-pipeline.yml to run debug unit tests. (#9722)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 27, "deletions": 5}}, {"node": {"oid": "31dc768e0760cb76114f1cf9e12524b4940ca0ad", "messageHeadline": "update ONNX Runtime Web CI to use same script for package versioning \u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 85, "deletions": 67}}, {"node": {"oid": "1541784f6c0e03a6defb8f562f7ae96368dd09ef", "messageHeadline": "[python api] align api with other language bindings' treatment of exp\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 98, "deletions": 80}}, {"node": {"oid": "517fff0a392f0316da64b9351deac7b7fb9c05ee", "messageHeadline": "Add opset 15 support for NNAPI/CoreML EPs (#9711)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "adf98feb2c24eb266fa51f88939b192f6ed30028", "messageHeadline": "ATenOp Support for BCEWithLogitsLoss (#9670)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 65, "deletions": 1}}, {"node": {"oid": "1b70a14c513794d84219141ebbcddf88b55d834f", "messageHeadline": "Remove usage of wstring_convert (#9251)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 1, "deletions": 5}}, {"node": {"oid": "3db2390dabb1c9c393b36b7dd04c79babf039a79", "messageHeadline": "QDQ tool modification (#9616)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 108, "deletions": 38}}, {"node": {"oid": "e65f284476a837aa5d38ab9e58e29d2f61044c6e", "messageHeadline": "[js/web] Support WebGL for ort format models in benchmarks (#9661)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 23, "deletions": 17}}, {"node": {"oid": "de018f58e8378ce57ad572d1ebba2c492595f3c4", "messageHeadline": "Update manylinux build scripts (#9701)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 570, "deletions": 419}}, {"node": {"oid": "bdc279a7ed3ff2a1e65c52f84013cb5405f5ee13", "messageHeadline": "Use the same allocator following Pytorch (#9697)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 15, "deletions": 4}}, {"node": {"oid": "229c9a4e1c2508fcac0dbe38418db1c9e9012833", "messageHeadline": "Added Trilu CUDA kernel. (#9633)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 412, "deletions": 48}}, {"node": {"oid": "6420530b3a71c84d66e1f79b27d16df952ca4fdd", "messageHeadline": "fix the mkl dependency for eager mode (#9702)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 14, "deletions": 3}}, {"node": {"oid": "a70ae244751a6d642a5f712eefdbf69151e9a86e", "messageHeadline": "Add QDQ::Selector::Select to use const GraphViewer instead of mutable\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 402, "deletions": 146}}, {"node": {"oid": "65590b049c0c1dcc1b0c9428e8c071559e652b4f", "messageHeadline": "Expose an API to query the CUDA compute stream to launch a custom ker\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 113, "deletions": 58}}, {"node": {"oid": "c579ebfbc3f07aa5636093ef033ff4187074a7d1", "messageHeadline": "change a for iteration (#9678)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "53afaefe3b82731a01c945b96e7a88862b91f54b", "messageHeadline": "Refactor Windows CI pipeline yaml files (#9672)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 375, "deletions": 311}}, {"node": {"oid": "24e35fba3217bf33b0e4064bc71d271a61938ba0", "messageHeadline": "Change TensorShape to typically not allocate heap memory (#9542)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 612, "deletions": 526}}, {"node": {"oid": "13e64f8ff7081e1466b02fbe824c7bee33e0d180", "messageHeadline": "Remove all warnings C4800: Implicit conversion from 'int32_t/int64_t'\u2026", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 42, "deletions": 40}}, {"node": {"oid": "d9c42acec1d0a794a7a40abf5e217bccfbdcce51", "messageHeadline": "Merged PR 6646876: Merge ORT Github master to DmlDev", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 5978, "deletions": 1517}}, {"node": {"oid": "7e207ba3be3620ffbb9f8b8fcffd15e911a744e3", "messageHeadline": "Use ORTMODULE_ONNX_OPSET_VERSION to modify the opset version in OrtMo\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 91, "deletions": 1}}, {"node": {"oid": "1151c661ebc7c50a7366d64d1fc31c7099d8ea65", "messageHeadline": "Add gi overload (#9690)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "f8ecf10cfb485af4df8051c8092a6e3e9dde587f", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/jeffbloo/Mer\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 5978, "deletions": 1517}}, {"node": {"oid": "c6fddb263fdff8cd25d28215645c7db58575138f", "messageHeadline": "Add Node.js binding support to packaging pipeline (#9577)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 670, "deletions": 34}}, {"node": {"oid": "1cbbafdbe0eddde94e0d77835ee56ff2aadf85f6", "messageHeadline": "Change the default value of onnxruntime_DISABLE_RTTI (#9674)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 7}}, {"node": {"oid": "3e5dbfd94f30ee328f3fc6b305f37d10e44a1d85", "messageHeadline": "Increase binary size limit and publish binary size data even if limit\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 13, "deletions": 5}}, {"node": {"oid": "dfe4d0a330c5b970217367d8fb8a77eebe4aca84", "messageHeadline": "Abjindal/eager windows ci pipeline (#9587)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 93, "deletions": 0}}, {"node": {"oid": "74dc48f61c07c144375693bc0d9e3dde13da5e02", "messageHeadline": "Remove python 3.6 from macos pipeline (#9602)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "ee167bd07886dc4e15386af2ad7bbf4a8024f0f4", "messageHeadline": "Optimize _TileKernel for non-memcpy case (#9648)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 96, "deletions": 5}}, {"node": {"oid": "a355bcbd7397be526efc9c54616ecaf4d7b05f34", "messageHeadline": "Clarify cgmanifest.json update process. (#9664)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 21, "deletions": 12}}, {"node": {"oid": "90de3c8e7c4158abb71fe26303bd27d552d88899", "messageHeadline": "Update MLOperatorAuthorPrivate.h (#9575)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "bbeceb7541860d99f58fe9b418e21a1a9dac3f83", "messageHeadline": "Support optional type in ORT (#8339)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2300, "deletions": 872}}, {"node": {"oid": "2fb03769bc5a8b53c9850d4562953ec1def8e4ab", "messageHeadline": "Updated OperatorUtility to avoid C2672 and C2783 (#9651)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "1128bf282d11c93665f62bc88477fd149396ca29", "messageHeadline": "Updated cpuid_info.cc to fix PVS-Studio error C1012 (#9581)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ddb4c058526a0bc6f866be7c297354ed65f0fbc5", "messageHeadline": "Save graph runtime optimizations for minimal build (#9508)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1912, "deletions": 261}}, {"node": {"oid": "71a1a7b4711b2cc42b528c7e980d42a5a0fc5cdc", "messageHeadline": "Enable building winml with --build_nuget (#9632)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 97, "deletions": 46}}, {"node": {"oid": "9cedb12c89c12ac37040b869a37533be69b990d0", "messageHeadline": "Add App Center e2e test to Android package (#9653)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 136, "deletions": 10}}, {"node": {"oid": "c8151b40372135da8d32f6313acb1f3b5f8a8be8", "messageHeadline": "Add percentile method for PTQ (#9342)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 140, "deletions": 15}}, {"node": {"oid": "2406a425a7678fada91e41c87a812c23973e6145", "messageHeadline": "Move the common part of Windows CPU CI pipeline to a template file (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 250, "deletions": 202}}, {"node": {"oid": "15867dcc082c74a9d308a5aa2af21777eed54aef", "messageHeadline": "enable prefast check with orttraining win pipeline (#9638)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "6ff02b04a842b0f71b410619cdcad47ba4cbd2d3", "messageHeadline": "Add Java build/test dependancies JUnit and Google Protobuf Java to cg\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "230099e482f1baf10c45e3dd53aafe0d55176855", "messageHeadline": "Make ORTModule serializable (#9634)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 223, "deletions": 41}}, {"node": {"oid": "3a1b4045c98b7b66f54c33ce86292e65dcb02101", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 2070, "deletions": 1163}}, {"node": {"oid": "51f84a0330921c1e0f76a996cc404a1a1ed90f2b", "messageHeadline": "support quantization of empty tensor (#9640)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 23, "deletions": 6}}, {"node": {"oid": "4644f8bc2c8fbcb6236ecb9132b6cacfd8e961da", "messageHeadline": "refactor profiling logics (#9567)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 5, "deletions": 6}}, {"node": {"oid": "19804876d9a6a529f50050c5913b5042213a07fb", "messageHeadline": "Updated data_types.cc to avoid warning C4946 (#9611)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "e11fde01797bd99d95504911546ebf567d85c795", "messageHeadline": "libonnxruntime_providers_rocm.so and libonnxruntime_providers_shared.\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 18, "deletions": 3}}, {"node": {"oid": "020ec9c0f57bc652cfb7806715e1f4fc08946986", "messageHeadline": "[DmlEp] Added missing include to DmlOperator.cpp to its own header (#\u2026", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "88f61a1b2df27222371c5c364f060a1e867fef6d", "messageHeadline": "[DmlEp] DmlEp acknowledges ORT_NO_EXCEPTIONS (#9622)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 898, "deletions": 654}}, {"node": {"oid": "f41396eff0bdcc89c9b2bc496af49cccb1862527", "messageHeadline": "Small addendum to #9624 (#9639)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 13, "deletions": 0}}, {"node": {"oid": "1240c595a9fc40b0d5d1351b9ef96772c9f590ad", "messageHeadline": "Fix a couple of bugs with how memory pattern planning is enabled/disa\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 14, "deletions": 2}}, {"node": {"oid": "c1112813e76689e01f66b3f92b11c8c424fa82fe", "messageHeadline": "Added missing _In_ to ort_apis.h to be consistent with its cc file (#\u2026", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5247247b7fb5f45f589b8e8baf1b7c0011145dd0", "messageHeadline": "Add Resize Handler for Transpose Optimizer and revise Layout Transfor\u2026", "author": {"name": "TomWildenhain-Microsoft", "email": "67606533+TomWildenhain-Microsoft@users.noreply.github.com", "user": {"login": "TomWildenhain-Microsoft"}}, "additions": 518, "deletions": 345}}, {"node": {"oid": "d203cdf132209aa20376f118a5dc59e01574f650", "messageHeadline": " Check output edges in QDQ (#9600)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 92, "deletions": 1}}, {"node": {"oid": "ccf1b8ad00aa8bd139e7db79fddb0967f9768e9b", "messageHeadline": "[DmlEp] Avoid warning C4495 (#9533)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "034f297884751576b3014f450a0e9651d211ed95", "messageHeadline": "[DmlEp] Avoided warning C4458: declaration of 'X' hides class member \u2026", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c315d1b3cdee1d5662e529d7fdce120eefa18193", "messageHeadline": "Always enable ORT format model loading. (#9586)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 121}}, {"node": {"oid": "5c56fa0def02471daae421330e239b688f45a9fc", "messageHeadline": "Miopen conv grad (#9574)", "author": {"name": "groenenboomj", "email": "groenenboomj@users.noreply.github.com", "user": {"login": "groenenboomj"}}, "additions": 462, "deletions": 7}}, {"node": {"oid": "a07da457d1bc39c5b3fe85566fedee5a99a46ba6", "messageHeadline": "Merged PR 6622182: add back change history that I lost with a squash \u2026", "author": {"name": "Brian Martin", "email": "brianma@microsoft.com", "user": {"login": "martinb35"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "6a2d709db48f17765d678208ac071c034d5a345a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into dmldev_temp", "author": {"name": "ISS Build Account", "email": "brianma@microsoft.com", "user": {"login": "martinb35"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "6c236fd9398a6d3172e93ef6127942edeb0bfb7d", "messageHeadline": "Merged PR 6622174: merge latest onnxruntime into dmldev", "author": {"name": "Brian Martin", "email": "brianma@microsoft.com", "user": {"login": "martinb35"}}, "additions": 87428, "deletions": 2824}}, {"node": {"oid": "79436a2d5bfbe79a5c900ab95f2926928603284f", "messageHeadline": "Avoided warning C5038 (#9543)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 33, "deletions": 31}}, {"node": {"oid": "a3e4d77b357ed36d9414029afc7f8cc1f863726f", "messageHeadline": "Replace wstring_convert with std::wstring (#9614)", "author": {"name": "Jingqiao Fu", "email": "jingf@microsoft.com", "user": null}, "additions": 5, "deletions": 3}}, {"node": {"oid": "f7774a91d61cd13675c8d320039804b0c0449b46", "messageHeadline": "Add api-ms-win-core-com-l1-1-0.dll, shlwapi.dll, oleaut32.dll to dela\u2026", "author": {"name": "Jingqiao Fu", "email": "jingf@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fa4658e8a9e009f27de77d0e2598bdad6de2a26e", "messageHeadline": "Move to XCode new build system if building on Mac using XCode (#9617)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 13, "deletions": 4}}, {"node": {"oid": "57491b6f9312c08b529bf02ed7c80613d1e7f88b", "messageHeadline": "Add App Center test for iOS package (#9605)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 83, "deletions": 13}}, {"node": {"oid": "b5f7bb7d103f97a82d2c1e50364b4598cf63c93b", "messageHeadline": "Update ONNX (#9462)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 49, "deletions": 25}}, {"node": {"oid": "7744cc1013a754f2323952bf17ef5a8f25e3b888", "messageHeadline": "[DmlEp] Make DmlEp compatible with Clang for EPIC (#9585)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 297, "deletions": 191}}, {"node": {"oid": "eb2612b5885d0036b374ae78308db2f64991ae79", "messageHeadline": "Remove netcoreapp2.1 target as it is EOL and out of support. Attempti\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "173e538b8018ae9f3684ea8055a80403e14c17b2", "messageHeadline": "Update mac-ios-packaging-pipeline.yml", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "cc73bcc243d10f9c49ceec328ad27a437a711e6b", "messageHeadline": "Suppress component governance component warnings for ios", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "1731f0080ae37176d729c455cc3bf17be9547674", "messageHeadline": "Update attention_cpu_base.h to suppress static analysis warning", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9c15c68ed41b2090e20bb178b66745d7093dab2d", "messageHeadline": "Enable fallback when forward fails due to non contiguous tensor (#9369)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 124, "deletions": 3}}, {"node": {"oid": "a01a3f2552a7a82b1264e0eb7c2ef38be065a1f5", "messageHeadline": "Add more statistics in transformer profiler (#9578)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 187, "deletions": 57}}, {"node": {"oid": "85874bb3150b9d9a7cce0ed3fa6500b49e385c2b", "messageHeadline": "embed layer fusion gpt2 (#9336)", "author": {"name": "Viswanath Boga", "email": "44417868+viboga@users.noreply.github.com", "user": {"login": "viboga"}}, "additions": 331, "deletions": 28}}, {"node": {"oid": "a555740708906c54dca4053d1446e2e4da1a002c", "messageHeadline": "Attention fusion: update uint8 tensor parsing for ONNX upgrade (#9564)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 18, "deletions": 2}}, {"node": {"oid": "17cf39a96499e7cd805a40b97939c2cf68d38c1f", "messageHeadline": "Clean up unnecessary codes in softmax and hardmax kernel (#9580)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 4, "deletions": 20}}, {"node": {"oid": "e8268c9a18ef4f4d8589ac0483b4d0f6f458e3cd", "messageHeadline": "Add Transpose Optimizer and modify nhwc optimizer to use it. (#9284)", "author": {"name": "TomWildenhain-Microsoft", "email": "67606533+TomWildenhain-Microsoft@users.noreply.github.com", "user": {"login": "TomWildenhain-Microsoft"}}, "additions": 6928, "deletions": 532}}, {"node": {"oid": "87b1fddd97e68892bfcf4ecdff1a2b031e01941e", "messageHeadline": "Add Linux/MacOS ARM64 support to nuget packaging pipeline (#9570)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 266, "deletions": 79}}, {"node": {"oid": "2d44bd525b8f7e44d7a91e04c0721d7a22146a80", "messageHeadline": "DML functions always returning a value (#9485)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 37, "deletions": 10}}, {"node": {"oid": "8dd4d207e27021f5aca9b70bad5c9ea68e4ad238", "messageHeadline": "Merged PR 6606929: RI 10/26 from github into fork #2", "author": {"name": "Sheil Kumar", "email": "sheilk@microsoft.com", "user": null}, "additions": 101, "deletions": 83}}, {"node": {"oid": "a2b3e6bb2338d9863486e3a6d4c3cc86cf123064", "messageHeadline": "Remove pointless assert. (#9571)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "4e76360261259fff3fc316b3889766ff4239c32f", "messageHeadline": "Prevent PySparseTensor form being garbage collected if we have an out\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 47, "deletions": 20}}, {"node": {"oid": "aa76520e60b0be9b2ff81de8db3861b2a1d7cc31", "messageHeadline": "Update macOS build agents to macOS 11 (#9562)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "5d5c03bcdce531b0c30bab0fac5fbaa8e0f38cb4", "messageHeadline": "Fix opset version change by not using copy of global constant (#9393)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 60, "deletions": 37}}, {"node": {"oid": "b5a652c578c0e4f66498531c5d894a2ea269eae4", "messageHeadline": "Add Xamarin support (#9436)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 78252, "deletions": 1666}}, {"node": {"oid": "12f216aab518fa92aa61234bd9963ef7a2de62c5", "messageHeadline": "Bug in DmlOperatorResize.cpp with m_inputDimensions (#9456)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9639eded4bc5ea191d47d22668909ef3658389f7", "messageHeadline": "Missing `#pragma once` in dml_provider_factory.h (#9457)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "1efdbff1a3742e23a9e37843e59ab9e6119e291e", "messageHeadline": "Fixed compiler error in Clang (for Win64) for ExecutionProvider (#9482)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0301f401eea2c1cae8f3b8da2c3e84cf1aa0a9ce", "messageHeadline": "Cleanup unnecessary opset_version arguments (#9558)", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 10, "deletions": 20}}, {"node": {"oid": "c79307e7b4bee3d7f4c3406a6559faaaf87c2fcb", "messageHeadline": "[js/web] support opset-13 of softmax (#9493)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 119, "deletions": 58}}, {"node": {"oid": "d079e0d48fc633b72d6e4110a1a677e72233077d", "messageHeadline": "Fixed Clang (on Windows) compiler error with `#pragma`'s (#9484)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "c54ad0dd0b7b7f2bafc64b49a36268acb59329e3", "messageHeadline": "POWER: Add Dgemm kernel for POWER processor (#9459)", "author": {"name": "RajalakshmiSR", "email": "58744256+RajalakshmiSR@users.noreply.github.com", "user": {"login": "RajalakshmiSR"}}, "additions": 533, "deletions": 5}}, {"node": {"oid": "90555bf96de509d633d67703fc9ee002f530b47b", "messageHeadline": "[node.js binding] enable CI for macOS arm64 (#9532)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 35, "deletions": 7}}, {"node": {"oid": "c1b0f924b7edfce48b5f80dbbcac4a4ae44f6b4c", "messageHeadline": "quantization tool better support operator when subgraph is enabled (#\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "33ef1d7700d84f980fbc9ce268d044c5882a43e5", "messageHeadline": "disable inner parallel for global avg pool as normally they are small\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 13, "deletions": 34}}, {"node": {"oid": "df7a5342a5bcf34d8f44dbfb77bbc6e52708dbda", "messageHeadline": "Upgrade com.diffplug.spotless to 5.17.0 (#9546)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f39821adbcf33cb883dc9ab2d26a5ae523e55123", "messageHeadline": "Fix a bug in CMakeLists.txt when handling NO RTTI (#9547)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 4}}, {"node": {"oid": "da15f5fc2f8a2ce0514db6a399f4c13cf352186c", "messageHeadline": "change cmake condition to prevent WCOS fom linking advapi32 (#9500)", "author": {"name": "Jingqiao Fu", "email": "jingf@microsoft.com", "user": null}, "additions": 1, "deletions": 4}}, {"node": {"oid": "542f1a9737d269e134eaff1c275d818dd10c0fa9", "messageHeadline": "Cleanup some whitespace and capitalization for set (#9504)", "author": {"name": "Stella Stamenova", "email": "stilis@microsoft.com", "user": {"login": "sstamenova"}}, "additions": 22, "deletions": 22}}, {"node": {"oid": "08dd8a0a02e6df89fd2ef4757b4392f6505686d4", "messageHeadline": "Merged PR 6604943: RI 10/26 from github into fork", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 17083, "deletions": 17176}}, {"node": {"oid": "a036cc6d4bbd87bc9598f0322ba1746c7c73745e", "messageHeadline": "Fixing bugs in ORT_NO_EXCEPTIONS (#9479)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "1aabba7120534af6d0576abcda4d6564f00033cb", "messageHeadline": "Avoided warning C4458: declaration of 'X' hides class member. (#9541)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "7dad90b200c428afb52d66c1f6ddb5081481fb41", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 17083, "deletions": 17176}}, {"node": {"oid": "f29057c7c0d775e027e0310dc95d818a67bf1738", "messageHeadline": "Added TanhGrad. (#9507)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 220, "deletions": 43}}, {"node": {"oid": "b125446f9c2ac6c527547ea428b11689977bc303", "messageHeadline": "Optimize python overhead of APEX amp (#9447)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 699, "deletions": 45}}, {"node": {"oid": "27ad20df2301cba38cec3800929cd6947197b926", "messageHeadline": "Add QDQ support of Resize to able to fuse it into a quantized Resize \u2026", "author": {"name": "Yi-Hong Lyu", "email": "yilyu@microsoft.com", "user": {"login": "yihonglyu"}}, "additions": 41, "deletions": 1}}, {"node": {"oid": "0270ff79512bf8c185fcc80928598d51446dc367", "messageHeadline": "Minor import fix (#9538)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f92b8e2ac8799ae63c4ed79777eea08a1040ecc0", "messageHeadline": "Clean up optional-lite references (#9534)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 81}}, {"node": {"oid": "bf4c3fa3d61b2ccdc792d96e4eabe7a6ee73733d", "messageHeadline": "[node.js binding] aggregate binaries for multiple platforms in single\u2026", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 132, "deletions": 719}}, {"node": {"oid": "fb4f7dbbb7ca1793c170253758653d94b035205c", "messageHeadline": "Call ATenOp for ReduceSum on ORTModule (#9471)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 122, "deletions": 35}}, {"node": {"oid": "651955d3c91be9a1cbb3ce29360758c8347016ad", "messageHeadline": "CUDA: Enable parallel compilation (#8974)", "author": {"name": "marcusfreisleben", "email": "85989692+marcusfreisleben@users.noreply.github.com", "user": {"login": "marcusfreisleben"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "39d1b9e1c176c29d9d6814a8625e675ab39f73dc", "messageHeadline": "Fix bug in Slice helper when dim value is zero (#9492)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 32, "deletions": 2}}, {"node": {"oid": "dbe1b57a71248946ade29c762817699fd1adc3ff", "messageHeadline": "Update thread_utils.cc", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a79d375d24b95c15954c2a8707def026d8583957", "messageHeadline": "Added fixes for Clang on Win64", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "9335cf102a3c65c837558a95b9b9cf012105303c", "messageHeadline": "Deleted duplicated \"core/graph/function.h\"", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 0, "deletions": 29}}, {"node": {"oid": "d6085044385332476b5e045246058b142a103660", "messageHeadline": "Don't use legacy mode for protobuf (#9498)", "author": {"name": "Stella Stamenova", "email": "stilis@microsoft.com", "user": {"login": "sstamenova"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "d83adaaf9fab62d57a84d88279d675667674a0bc", "messageHeadline": "Remove optional-lite (#9424)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 19}}, {"node": {"oid": "3ed8ade675bb2debafbdc0870733d1945913ec04", "messageHeadline": "Use SafeInt for malloc related computation (#9503)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "beddbdec5a100f1b94ae68d0c56d6c229fabb33b", "messageHeadline": "Fix PythonOp exporter (#9318)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "5adf17584751b055239ad98e8d0b28bb48457363", "messageHeadline": "pad shape 0 is not allowed in edge mode to comply with latest numpy (\u2026", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 32, "deletions": 8}}, {"node": {"oid": "d2d480a0db665dc9e69c8444a4f7475966eaeda8", "messageHeadline": "Allow None As Autograd Context (#9315)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 78, "deletions": 15}}, {"node": {"oid": "b64b2d48f302361cd490c0cb3a801a1e1ebdf66c", "messageHeadline": "Move iOS e2e test to XCUITest  (#9422)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 75, "deletions": 105}}, {"node": {"oid": "7f2f56633c7f6cf1beac967346b82a565dc6dff0", "messageHeadline": "Fixed implicit conversion warnings (#9481)", "author": {"name": "Gin\u00e9s Hidalgo", "email": "gineshidalgo99@gmail.com", "user": {"login": "gineshidalgo99"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "49b66c748604f56e0b4de1db657fecbcf963b0fa", "messageHeadline": "NFC: Normalize whitespace around if statements in CMakeLists.txt (#9464)", "author": {"name": "Stella Stamenova", "email": "stilis@microsoft.com", "user": {"login": "sstamenova"}}, "additions": 136, "deletions": 136}}, {"node": {"oid": "ca7116ca3eae0909919c4d5402f80a99ae21c615", "messageHeadline": "CUDA EP's ResizeImpl now uses functors, hipify for ROCm EP (#9466)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 198, "deletions": 211}}, {"node": {"oid": "66ceb6926df6c7a81dc1f241c6bf59cfe58d3a72", "messageHeadline": "rehipify ROCm EP files under orttraining (#9443)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 126, "deletions": 243}}, {"node": {"oid": "ff23b9ff55b7f90694a2e6abd64ec69afaa3c820", "messageHeadline": "Avoid cudaStreamSync at the end of Forward/Backward (#9470)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 19, "deletions": 13}}, {"node": {"oid": "5797bd6db3d7304b7eb216a0158d6a1d437adc4a", "messageHeadline": "Remove one unnecessary deepcopy in unflatten_user_output (#9353)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 17, "deletions": 15}}, {"node": {"oid": "4028e51e7e6421fdbeca5f4e4ccd8b4f790d0fd5", "messageHeadline": "Update the compatibility of ONNX Runtime Web (#9444)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1249c7c29e8d2d9cd5a341cc6be9df179c91b11c", "messageHeadline": "Resolve issue when running Yolov4 on DNNL EP (#9355)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 56, "deletions": 40}}, {"node": {"oid": "9fc53df33a3dd7ad20e061b610e18ea8c5b795a4", "messageHeadline": "Only add aliasing to targets if the corresponding package was found (\u2026", "author": {"name": "Stella Stamenova", "email": "stilis@microsoft.com", "user": {"login": "sstamenova"}}, "additions": 9, "deletions": 8}}, {"node": {"oid": "f1123c2fb31268b2e885dc7b1e5d4512674492ae", "messageHeadline": "Fix whitespace and style in concat.cc (#9452)", "author": {"name": "Nick Kreeger", "email": "nick.kreeger@gmail.com", "user": {"login": "nkreeger"}}, "additions": 29, "deletions": 26}}, {"node": {"oid": "89a22fb64177ec62527eb7d00cb8321492f5a6ef", "messageHeadline": "Add TopK to ROCm EP (#9391)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 20, "deletions": 10}}, {"node": {"oid": "f8acc6d0e814493eac0054f92259a3dc10f9ddd2", "messageHeadline": "Add NonMaxSuppression and RoiAlign to ROCm EP (#9394)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 14, "deletions": 13}}, {"node": {"oid": "c33391329a7cb85b2ac15257e0355d3f175cfbdb", "messageHeadline": "Add QuantizeLinear and DequantizeLinear to ROCm EP (#9401)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 4, "deletions": 8}}, {"node": {"oid": "406f1629c11464053e5af599042f111991d57281", "messageHeadline": "Remove Featurizers code (#9300)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 9122}}, {"node": {"oid": "e983f371211fd21bcfa9d25ec1a42f4fd92b648d", "messageHeadline": "Bifurcation detector for aggressive decoding (#9432)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 287, "deletions": 8}}, {"node": {"oid": "20eaed43e513633ce46d9e7d11db3db4d0d77da4", "messageHeadline": "Ignore all string inputs to ORTModule AB#1310803 (#9344)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 46, "deletions": 31}}, {"node": {"oid": "4698b73725afd2aa207827dfff7f0ce30d1c4188", "messageHeadline": "Fix output shape description of Attention op's schema (#9406)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3873885316196e4ff6e638dffa4af886c8ae5fe2", "messageHeadline": "add missing atomic include (#9440)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "52c53e396d1d6a104274ecb49be13b09e93057bf", "messageHeadline": "hipify tensor/gather_nd_impl.cu (#9392)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 0, "deletions": 118}}, {"node": {"oid": "a2ba923ac778998471be105c6889954dfd5c40d1", "messageHeadline": "hipify fast_divmod.h (#9400)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 1, "deletions": 61}}, {"node": {"oid": "a8e2e8d76a5f33eb9894ce8adba169f807ac00e3", "messageHeadline": "hipify tensor/transpose.cc and tensor/transpose.h (#9397)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 3, "deletions": 284}}, {"node": {"oid": "757bc667204b1b9a2d5993c2a914aaf643eaa463", "messageHeadline": "Set cuda version to be None instead of an empty string (#9435)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "e22920d954909b6c940473fd2df876076893c36b", "messageHeadline": "Update ORTTraiing frontend codeowner (#9427)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "da3dd398c5e03b1309182af1979f381961d07715", "messageHeadline": "Kernels for QLinearConv with symmetrically quantized filter (#9323)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 4879, "deletions": 157}}, {"node": {"oid": "5da4e07daa02bbfea1edece08ba52a76cec13d13", "messageHeadline": "Make FusedAdam mathematically equivalent to Transformers AdamW (#9343)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 76, "deletions": 82}}, {"node": {"oid": "5b65f1cb44eee6d82b9ba3b91a43b5b4a74ccae8", "messageHeadline": "fixes SDL Native Rules warning in Node.js binding CI (#9402)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 210, "deletions": 72}}, {"node": {"oid": "f60e60302206ec7d54e48fabe02674198067230e", "messageHeadline": "Add support for DmlExecutionProvider for transformer profiler tool (#\u2026", "author": {"name": "Jingqiao Fu", "email": "jingf@microsoft.com", "user": null}, "additions": 32, "deletions": 10}}, {"node": {"oid": "0824207c0f44accf4a7949616dfc24c665eb0345", "messageHeadline": "Add Dev Guide to transformer optimizer (#9329)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 53, "deletions": 0}}, {"node": {"oid": "6ecb990fae723510248bfa062c1ec2032b29e3aa", "messageHeadline": "Update win-ci-pipeline.yml", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b130a7b7156ebf974c54688a9f159d212850043e", "messageHeadline": "fix MSVC micro benchmark build warnings (#9373)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 14, "deletions": 2}}, {"node": {"oid": "59dfab59dc37949f7d99c962a3706b6d88c3ee12", "messageHeadline": "Fix integer overflow for large step for Slice OP (#9376)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 47, "deletions": 24}}, {"node": {"oid": "901c7de918e63a0c7f77cc0ecf3b4e12c969ad5d", "messageHeadline": "[js/web] remove webgl from default fallback list (#9374)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 11, "deletions": 8}}, {"node": {"oid": "f05c285a58a98181773ce32cf3dcff64b54cb141", "messageHeadline": "Exception when duplicated autograd.Function name detected (#9351)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 28, "deletions": 8}}, {"node": {"oid": "74eaaad768e0a5ad5e65b818605383c73dece509", "messageHeadline": "[js/web] Support opset-13 for squeeze, unsqueeze, maxpool, pad, cast \u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 273, "deletions": 107}}, {"node": {"oid": "c8789d3047d4a23cafeb4c0950ad6be5c6b3c37c", "messageHeadline": "[ROCm] static re-hipify of CUDA EP to ROCm EP, now a shared provider \u2026", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 2434, "deletions": 1686}}, {"node": {"oid": "87e726d1a0877a5ec98ac67154ef8e693e717d7c", "messageHeadline": "Abjindal/merge eager with external custom ops (#8986)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "23700a15a095515a38408f565bf97a526fa7a311", "messageHeadline": "Abjindal/eager windows build (#9326)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 53, "deletions": 19}}, {"node": {"oid": "3e879aab6bf5f6e902151e9d14c4fe76f6792e1f", "messageHeadline": "work around ucx in rocm ci Dockerfile (#9360)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "11f0081c1e822444452d6e560940914fbdba7e0f", "messageHeadline": "Remove tensorflow, tf2onnx from the list of dependencies for the docu\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 126, "deletions": 79}}, {"node": {"oid": "22e3f8bf549e66ae2f0b6baa64ca01656e2865ad", "messageHeadline": "Refactor TrainingManager.forward (#9354)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 129, "deletions": 124}}, {"node": {"oid": "851554536ca8185b3413ee57449ea5ac93370193", "messageHeadline": "[DML EP] ConstantsOfShape - Empty Output and EinSum - Optional Parame\u2026", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 31, "deletions": 4}}, {"node": {"oid": "5ee47e3ffad950e9837bdbe5f2ac36b69c5db49f", "messageHeadline": "legacy_megatron-lm/deepspeed_ZERO1&2 FP16_Optimizer wrapper (#9184)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 460, "deletions": 0}}, {"node": {"oid": "4771256be3cf2992cc87797e8b215f1b55c91d7b", "messageHeadline": "fix to avoid quantizing attention with varied q,k,v sizes (#9357)", "author": {"name": "Viswanath Boga", "email": "44417868+viboga@users.noreply.github.com", "user": {"login": "viboga"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "ba0cca96f033949530f1b2cc493133a6a14dec5c", "messageHeadline": "Hooked up eager logging to ORT default logger. (#9340)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 86, "deletions": 147}}, {"node": {"oid": "905fe365998c4b1288d0814f286e2fe7ee05d2fe", "messageHeadline": "Add Conv and ConvTrans to ROCm EP (#9338)", "author": {"name": "groenenboomj", "email": "groenenboomj@users.noreply.github.com", "user": {"login": "groenenboomj"}}, "additions": 883, "deletions": 12}}, {"node": {"oid": "bccd09c6884088c67b68d95a85b232a13f860d91", "messageHeadline": "Serizalize model only once to reduce backend preparation overhead (#8\u2026", "author": {"name": "Arthur Meyre", "email": "arthur.meyre@zama.ai", "user": {"login": "IceTDrinker"}}, "additions": 13, "deletions": 2}}, {"node": {"oid": "e8ba5145ceabad258ab735da41acf671a0bfa49a", "messageHeadline": "Add  Transpose, Reshape, Pow and  LeakyRelu ops to DNNL  execution pr\u2026", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 727, "deletions": 143}}, {"node": {"oid": "1527af3e308afae3a05cb0bb4742ec2ba0ab6a22", "messageHeadline": "[js/web] deduplicate test cases between opsets (#9327)", "author": {"name": "Yulong Wang", "email": "7679871+fs-eire@users.noreply.github.com", "user": {"login": "fs-eire"}}, "additions": 73, "deletions": 30}}, {"node": {"oid": "fb31701f7e043d88216345521782771a847392ce", "messageHeadline": "Fix bug in determining default slice axes (#9328)", "author": {"name": "TomWildenhain-Microsoft", "email": "67606533+TomWildenhain-Microsoft@users.noreply.github.com", "user": {"login": "TomWildenhain-Microsoft"}}, "additions": 13, "deletions": 2}}, {"node": {"oid": "510b747821dc0f7ce32c0bfecdd5c616fab95b85", "messageHeadline": "w (#9319)", "author": {"name": "Moshe David", "email": "mosdav165@gmail.com", "user": {"login": "mosdav"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f0bc35c4bab68345fc21038d54d415b3bdfe5bc8", "messageHeadline": "fix a hardcode type (#9337)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d5c5c4fa500eaabc5e48775437c10b608a0a66e7", "messageHeadline": "Handle implicit subgraph inputs required on different devices in Memc\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 200, "deletions": 7}}, {"node": {"oid": "48737091c0ea2d7b34ce48d0e67f276776eba3d9", "messageHeadline": "resolve the provider options before create training session in orttra\u2026", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 288, "deletions": 87}}, {"node": {"oid": "52c021d1f3e7d7e2bae15aabd748ac3c9566f6b1", "messageHeadline": "Fix export of aten op for Max and Avg Pool 2D (#9330)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 40, "deletions": 0}}, {"node": {"oid": "f9cf62912a317d6d13e074796c847b47739ecc5d", "messageHeadline": "Add same_shape case for BiasDropout (#9188)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 230, "deletions": 55}}, {"node": {"oid": "2f1204a5d5d1d017508c0ec1be47755901458b0c", "messageHeadline": "[js/web] Enable wasm profiling and preserve function names in profili\u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "787dcb7dbc2b65563c6d3ecd3f65ab5a2e5d8172", "messageHeadline": "Support extra addition before softmax in attention cuda kernel (#9205)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 84, "deletions": 62}}, {"node": {"oid": "03276527b3b77db97f8a73fbce449ca0e0bcee50", "messageHeadline": "Fix typing error (#9316)", "author": {"name": "Jiaxu Dong", "email": "dongjiaxus@163.com", "user": {"login": "xusworld"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "79e736ed258ac5b5bb8a4420a4d0449bc13339c6", "messageHeadline": "Make onnxruntime::Status nodiscard (#9279)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1343, "deletions": 1349}}, {"node": {"oid": "da56f01ac20ac7e1793a3847ec6c8d3111b5e7a6", "messageHeadline": "Fix bug in ReduceSum with noop_with_empty_axes (#9301)", "author": {"name": "TomWildenhain-Microsoft", "email": "67606533+TomWildenhain-Microsoft@users.noreply.github.com", "user": {"login": "TomWildenhain-Microsoft"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "7b61bca6dfd6c29f0969fbab3c16698edf638295", "messageHeadline": "Fix inclusive sum overlfow when applied on int8_t buffer in Compress \u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 53, "deletions": 10}}, {"node": {"oid": "29379db4326888b88f21c4e267400ec7f2d86462", "messageHeadline": "Added SigmoidGrad schema and kernels. (#9244)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 196, "deletions": 10}}, {"node": {"oid": "cd65a8089edb9e5cb2fd60de3a4311c3dd362a35", "messageHeadline": "Optimize Variadic Elementwise Ops (#9186)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 98, "deletions": 48}}, {"node": {"oid": "5f5f28bf14ff29d9140a6b4ebc320da0f4d6713a", "messageHeadline": "Fix bug in allocation planner while planning location for initializer\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 25, "deletions": 7}}, {"node": {"oid": "68601fc296f2c20e7d9fe7156ebd824f8d585a13", "messageHeadline": "error handling ffor eager mode's data transfer (#9261)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "70cf61fa84bf694d9858fb370a8d383d27cd0219", "messageHeadline": "disable bart-l for now (#9305)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 27, "deletions": 27}}, {"node": {"oid": "72c4cea9e66f0c02933291ad377b7d2e162962cb", "messageHeadline": "[OpenVINO-EP] V3.2 Release (#9232)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 134, "deletions": 14}}, {"node": {"oid": "7166586d7ebcb0169980b188615b8044f06de7f4", "messageHeadline": "Enable SkipCheck by default (#9215)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 66, "deletions": 12}}, {"node": {"oid": "88d5023885ecfde70a5947a3247ab430f9270fb8", "messageHeadline": "[js/web] always use new data dir for ort web E2E karma tests (#9303)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "c002dc86a364852859ca9642698fcfc5edf22c9d", "messageHeadline": "set mpi group  init flag after add group (#9293)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "4f4875b0e8c21c92fe7af05151acf36ff92828c8", "messageHeadline": "Add   \"workspace:   clean: all\" to anybuild build yaml file", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "e2b1852eecc82b92daeae27ec0692d4197b1bd73", "messageHeadline": "Build: respect onnxruntime_PREFER_SYSTEM_LIB for more things (#9181)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 24, "deletions": 7}}, {"node": {"oid": "52d067402a09efcf8d47c7e34c4479b3f3168ab9", "messageHeadline": "Fix all-or-nothing fallback for bad ORTModule init (#9277)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 47, "deletions": 35}}, {"node": {"oid": "510b58c8774b27847b542a46567e4c4bcf87ca81", "messageHeadline": "Increase AMD CI pipeline timeout to 120 min (#9280)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 274, "deletions": 271}}, {"node": {"oid": "334980e016fd6531a09467efad05009bc9aa1a6d", "messageHeadline": "Delete nocontribops pipelines", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 735}}, {"node": {"oid": "bcdb411c8d177d5b931e21b199cce24a31a5ab98", "messageHeadline": "Implement FusedAdam for ORT adapted from DeepSpeed (#9266)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 637, "deletions": 1}}, {"node": {"oid": "a4d53c4ab5d097497da16fe3e788d034d2c372a5", "messageHeadline": "fix training distributed ci failure (#9273)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 14, "deletions": 3}}, {"node": {"oid": "35c2102cfa85446f4a47ebc3d172c39c744f506e", "messageHeadline": "Fixes for GatherND, Multinomial (#9143)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 122, "deletions": 13}}, {"node": {"oid": "0b77c9ca7ce34e160222c2c165ea4e7daf8d56a0", "messageHeadline": "Cleanup function definitions of contrib ops (#9265)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 182, "deletions": 171}}, {"node": {"oid": "ccf8ef67603b4cb6d0292c9429e74703f731954e", "messageHeadline": "Merged PR 6531168: Unsquash merge on DmlDev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "21e29cf17b64c381f1192aed544488e58467cade", "messageHeadline": "Merge remote-tracking branch 'origin/DmlDev' into user/ticastro/fix-ri", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "6e2f66ee9c3214acf5de9b54484a9d69a804f0cc", "messageHeadline": "Allow custom exporter args + bug fix (#9242)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 97, "deletions": 11}}, {"node": {"oid": "67ff339df7f3b6a893bc0380e835ba921d002ef4", "messageHeadline": "fixed a profiler.py bug (#9231)", "author": {"name": "Jingqiao Fu", "email": "jingf@microsoft.com", "user": null}, "additions": 10, "deletions": 8}}, {"node": {"oid": "113edbda64ca32e5d445b90cad9ecbfa2ea2bba7", "messageHeadline": "Add bf16 specialization for IsDataType (#9254)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "8f6fd014e461ec9096a537365099c508cd46aacf", "messageHeadline": "Force Windows AI NuGet pipeline to use Windows SDK 19041 (#9255)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9fe09cb72a3e332ea7cfc7449978d505023a7d8f", "messageHeadline": "Update dockerfile readme (#9241)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 17, "deletions": 317}}, {"node": {"oid": "20b9390d1d13ea0ead63dc17282f6cd24cb109a0", "messageHeadline": "Merged PR 6524907: Fix merge conflicts from public ORT to WindowsAI ORT", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 64071, "deletions": 39607}}, {"node": {"oid": "335283d299d6b6cf0fb4cfec45ff895d15f878da", "messageHeadline": "Fix merge conflicts, merge from public onnxruntime", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 64071, "deletions": 39607}}, {"node": {"oid": "11a391a88ff9e59136f5cee82c647dc1b4a63be0", "messageHeadline": "Port ARM64x support (#9230)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 25, "deletions": 0}}, {"node": {"oid": "60bbdf14035014d94d525c402d3d6404ac32ad3d", "messageHeadline": "Remove unused NodeArgs in Graph::Resolve (#9213)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 82, "deletions": 26}}, {"node": {"oid": "8adb9ab85a996e40db045c8b2d9f60a065f412d4", "messageHeadline": "fix CodeQL warning for path-injection (#9243)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 40, "deletions": 43}}, {"node": {"oid": "45399d5ace262e9d4cad07490d6c483670e12729", "messageHeadline": "Remove TORCH_WARN to avoid torch string related operations that take \u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "be4d8874398db47da03f3e3fc2fa12c66a100ae1", "messageHeadline": "Fix ONNX exporter call with latest API for ORTrainer (#9228)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "448325b25455242e97c1784c35ded32569d1c3e7", "messageHeadline": "[js/web] name ort web for consistency (#9240)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "c23a216900493f252244e64b014eeb52d40fa395", "messageHeadline": "MLAS: fix AVXVNNI+Linux qgemv kernel (#9234)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e2d779246af4b98e7af10d543b573e6830f72a87", "messageHeadline": "[wasm] remove deprecated prefix 'EXTRA_' in emcc flags (#9211)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "c6cb49c5a10dc4d4cb018c25b2a84359baf4ffae", "messageHeadline": "DirectML.dll load fails when executable path contains Non-English cha\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 12, "deletions": 17}}, {"node": {"oid": "634bb5ede0ab0e051d3ce7343812a13e7b336d26", "messageHeadline": "fix CodeQL warning 'Remote property injection' (#9224)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "8c57d51928de606b1027850b9df0cf32073a4b6e", "messageHeadline": "support WebAssembly SIMD for qgemm (#9191)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 531, "deletions": 7}}, {"node": {"oid": "e79be39081ab3f9eb4a383427ffdb41e5ec717fe", "messageHeadline": "LayerNormGrad function body and LayerNorm inference/body fix (#9160)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 155, "deletions": 10}}, {"node": {"oid": "e1b84eefcce51fead0b70428a274176c4149cb75", "messageHeadline": "Revert \"Revert \"linux trt package pipeline (#7537)\"\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 66, "deletions": 0}}, {"node": {"oid": "5326397a6a7d761125a61b941cd0a742672c6fa0", "messageHeadline": "[iOS] Facilitate usage of pods with custom builds (#9216)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 179, "deletions": 41}}, {"node": {"oid": "ceb51dda4a6c48fced3e9d3df333b1702b1f83db", "messageHeadline": "Support external torch cpp extensions on ORTModule (#9223)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 138, "deletions": 118}}, {"node": {"oid": "ffca0b777be6d1522cf6665a4695ddbfd1585a77", "messageHeadline": "Patching cuda profiler with enhancements (#9214)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 39, "deletions": 21}}, {"node": {"oid": "beab1ef1bbd2b9c930e5ed005ae0ef2376a9d79b", "messageHeadline": "Merged PR 6516718: [DML EP] Direct INT64 support for indices tensor f\u2026", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal@microsoft.com", "user": {"login": "sumitsays"}}, "additions": 11, "deletions": 10}}, {"node": {"oid": "dae5b1d4ac2870cbf2ba5af19421a1dce12b91b1", "messageHeadline": "Merged PR 6515621: Autopilot build ORT clamp_cast on *64-bit release", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 40, "deletions": 19}}, {"node": {"oid": "4a1b386f7cd207e580add1ef0788a3c973f633f1", "messageHeadline": "#9182 removed the `--is_store_build` option but one place where that \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "278928a102e96d64974299f1a2e4402f3c6a4d1d", "messageHeadline": "Added a test case for python gradient builder. (#9207)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 34, "deletions": 0}}, {"node": {"oid": "4f10024868e57ec16bb4c45e0fe576c122bae7d9", "messageHeadline": "Fix shape inference issue in Gather op (#9147)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 31, "deletions": 9}}, {"node": {"oid": "39c2d76c4e6c93c05b272cd17c87b07f17f59a7d", "messageHeadline": "Merged PR 6512710: Code clean up in DmlOperatorPadding", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal@microsoft.com", "user": {"login": "sumitsays"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "8bcd076f3d2cc3db062344d081500528607f8019", "messageHeadline": "Merged PR 6511109: Register ONNX operators for int64 in DML EP", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal@microsoft.com", "user": {"login": "sumitsays"}}, "additions": 803, "deletions": 56}}, {"node": {"oid": "b6060058588ae3757d1b28f423e0ad5e2bfdaa46", "messageHeadline": "Revert \"linux trt package pipeline (#7537)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 66}}, {"node": {"oid": "058108bef9769dd2768d0a37f06ce5e03d0aa8b5", "messageHeadline": "Execution Provider Profiler (#8406)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 434, "deletions": 72}}, {"node": {"oid": "6f580f07de81e126b876cf81757ea342b8835145", "messageHeadline": "Switch AMD CI pipeline to use environment image from onnxruntimecibui\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "1104e8d3e5cd9c1ce1bceb7c2700de72548807bf", "messageHeadline": "Linux Anybuild build pipeline (#9091)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 148, "deletions": 0}}, {"node": {"oid": "d3f859fe3063bdfc4b2fdadbd46985b3742b9aec", "messageHeadline": "Dropout Vectorized Kernel  (#9157)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 291, "deletions": 67}}, {"node": {"oid": "1b0816859f01b978131d9d45beae4b7e9244e651", "messageHeadline": "Only wrap sub-modules which can be wrapped as ORTModule (#9021)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 440, "deletions": 0}}, {"node": {"oid": "1a716871029bbcec7a3a1ec2778de3adcd4a5938", "messageHeadline": "Add QDQ for output of node (#9134)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 51, "deletions": 28}}, {"node": {"oid": "f16bb37fae410e490d4cc4ca30440385dd21e8d0", "messageHeadline": "make uwp store apps link to statically-linked crt desktop builds (#9182)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 36, "deletions": 209}}, {"node": {"oid": "c30cc9190a956b9d3a24639681699b8404ad36e7", "messageHeadline": "Change the agent pool for orttraining-distributed pipeline (#9179)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9fda95fec9925445d85d79ffab54c9b35542e029", "messageHeadline": "Change keepdims of ReduceMax/ReduceMin to always 1 when using quatiza\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 10, "deletions": 16}}, {"node": {"oid": "fd91bf91c943ac11f1963ec12a558a45372b084e", "messageHeadline": "Print full stacktrace exception when exporter fails (#9169)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 96, "deletions": 84}}, {"node": {"oid": "39dc6ea8a32d7dcdbeea45967452118ab16dcefc", "messageHeadline": "Fix to_dlpack Failure on PyTorch-1.10 (#9151)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 36, "deletions": 18}}, {"node": {"oid": "0888c6cc5926ac5debf85b105264933784591b82", "messageHeadline": "fix bug introduced by PR9130 (#9166)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "ddafe50199f79247234f2f5954fea6d94d06d875", "messageHeadline": "Fix Android build break after Virtual Environment update to 20210919 \u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "bde16eea684a12f2c9817539b6199108d3ff0f29", "messageHeadline": "Cleanup USE_TENSORRT macro (#8593)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 109, "deletions": 35}}, {"node": {"oid": "6e83392ff134f450aa5b8798ccbb138ec32b5379", "messageHeadline": "Bump up TVM version to avoid conflict with existing one (#9159)", "author": {"name": "ke1337", "email": "22626095+ke1337@users.noreply.github.com", "user": {"login": "ke1337"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4934455ab63d41f102295470f6a5e03b7677f7f2", "messageHeadline": "Bumping up to 1.10 (#9006)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 21, "deletions": 15}}, {"node": {"oid": "4e5bc8365b882563088542f877d4a83f4c08a958", "messageHeadline": "Add Paddle2ONNX to Versioning.md (#9067)", "author": {"name": "Jason", "email": "928090362@qq.com", "user": {"login": "jiangjiajun"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "267fb898e33c263c6337168556546aa5d6bc3632", "messageHeadline": "Added code to support Softmaxgrad for DNNL EP (#9022)", "author": {"name": "chethanpk", "email": "63478277+chethanpk@users.noreply.github.com", "user": {"login": "chethanpk"}}, "additions": 91, "deletions": 3}}, {"node": {"oid": "675ea9b3055fd3065ad20f78e9e4e87cb5762ac0", "messageHeadline": "Merged PR 6486412: Register ORT DML ONNX operators for int64", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 90, "deletions": 84}}, {"node": {"oid": "1db21da5aeb93a4b08a92a40566a0c630dc2641b", "messageHeadline": "Replaced onnx build with pypi installation (#9139)", "author": {"name": "sfatimar", "email": "sahar.fatima@intel.com", "user": {"login": "sfatimar"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "153767bab42cb5b51296c8fb269be6ccdeacf99d", "messageHeadline": "Add internal determinism flag configuration for ORTModule (#9074)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 213, "deletions": 117}}, {"node": {"oid": "b175f98dccb344ef6646f2b3d3731c47629039f8", "messageHeadline": "Do not generate nuget symbol packages on Linux (#9131)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "4df94a631d49592243f3a22a358471ebef3a80d0", "messageHeadline": "[NNAPI EP] Fix MaxPool error using uint8  (#9129)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 12, "deletions": 9}}, {"node": {"oid": "f7dedc90029cc0647b203a3407e2dee0be862e65", "messageHeadline": "Fix default initialization value in C API header (#9126)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 40, "deletions": 12}}, {"node": {"oid": "02b921373609a56493496d16f3b8770da69872e4", "messageHeadline": "Fix a bug for Openvino Python binding (#9130)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "dae37dc946f97c6ca2b30a801be868bbd52c6ed1", "messageHeadline": "Fix S360 issue by using \"use strict\" for javascript code. (#9128)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "49b329e26676e4a8f831e6829d4f893f78169f71", "messageHeadline": "[js/api] add typedoc and revise comments (#9077)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 247, "deletions": 14}}, {"node": {"oid": "750e2e0481579d877190fc8bb9b70adaf7d29347", "messageHeadline": "[js/web] check session ID in releaseSession() (#9105)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5e197698b8964875bb9b979070c325b51bb0228a", "messageHeadline": "Fuse attention for BERT without num_heads, hidden_size (#9108)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 56, "deletions": 25}}, {"node": {"oid": "47888392abd6b8065fe7b73863df99b6bdd7a07c", "messageHeadline": "Fix nightly CI pipeline to generate ROCm 4.2 wheels and add ROCm 4.3.\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 293, "deletions": 53}}, {"node": {"oid": "23e9c0a7f1b3005635d69cbe44c45730da851116", "messageHeadline": "move quantization examples to onnxruntime-inference-examples (#9107)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 3658}}, {"node": {"oid": "6ae5f7a24478bf34b402db3a176020024b5d0567", "messageHeadline": "C API Docs - Add build instructions (#9106)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 22, "deletions": 7}}, {"node": {"oid": "b876e5675badf515bb9a3948add3b7010a14c3c0", "messageHeadline": "C API Enum Name Fixes (#9092)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 25, "deletions": 25}}, {"node": {"oid": "0183492931816d0d6fc414dedbf1c164ac4d8630", "messageHeadline": "simplify Web CI pipeline (#9094)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 390}}, {"node": {"oid": "438175cb34d91ecc2f4d65c7c3092c3d734ecbd8", "messageHeadline": "Build shared host protoc in iOS CI pipelines (#9087)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 94, "deletions": 15}}, {"node": {"oid": "7fc28cd539850738c2127a2d37abc2a68a194d92", "messageHeadline": "[OpenVINO-EP] UEP v3.1 Release with OpenVINO 2021.4.1 (#9081)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 10, "deletions": 11}}, {"node": {"oid": "280e79463a3e7a4694682e955e9edc09ba2bea1d", "messageHeadline": "FIll in more documentation (#9088)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 244, "deletions": 78}}, {"node": {"oid": "26509465f0aa0a2e91a16d00dd6f3b36fcf2ace5", "messageHeadline": "Add default C++ initialization to OrtCUDAProviderOptions (#9064)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 32, "deletions": 76}}, {"node": {"oid": "12515552d1c89121f356862716c82d7cb9c9ac8e", "messageHeadline": "Remove cpuinfo from WCOS builds (#9076)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "498461485fa61af7e47cb34db1a48966799b47a1", "messageHeadline": "optimize WebAssembly CI pipeline (#9058)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 765, "deletions": 264}}, {"node": {"oid": "4828d2ebb19e10f86a1b4c431f073c70c70d7768", "messageHeadline": "MLAS: port aarch64 sgemv kernel to Windows ARM64 (#9071)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 318, "deletions": 12}}, {"node": {"oid": "bee5c2658048d0ff92dec5b88e6ec2ff8dcfec66", "messageHeadline": "Add CPU_ONLY runtime option to NNAPI EP (#9066)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 23, "deletions": 4}}, {"node": {"oid": "e758870b181ae63a10e3d282807459f4bb6d7836", "messageHeadline": "Upgrade ROCm CI pipeline for ROCm 4.3.1 and permit run inside contain\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 682, "deletions": 60}}, {"node": {"oid": "a05e32803ab5bb9dd69454d93ae4aca3999de6da", "messageHeadline": "Fixing MORE mlas unittest failures in POWER (#8673)", "author": {"name": "austinpagan", "email": "fossum@us.ibm.com", "user": {"login": "austinpagan"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "273494ee9e3a36c849d262ccce5f0dd28cb871f5", "messageHeadline": "Ensure ms-experimental domain Audio Ops build in mac pipeline  (#8857)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "98ac341c5b54dd6330fa4653a366ef6e5a00a996", "messageHeadline": "Filter nones from ctx saved tensors (#9063)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "4930320647143375397474919e0e0028b29ae90a", "messageHeadline": "Delete linux-pytorch-custom-ops-ci-pipeline.yml (#9023)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 32}}, {"node": {"oid": "0270ab17c56e33410dd223415472b8e0b616dc58", "messageHeadline": "Set onnxruntime_DISABLE_RTTI to default OFF (#9049)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "32366fea024c19fdf4d63cc73884103d50836584", "messageHeadline": "[Objective-C API] WIgnore clang documentation warnings from C/C++ hea\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 38, "deletions": 21}}, {"node": {"oid": "3ec3e9f70534ca13823add8bfc75722f07d67253", "messageHeadline": "Add t-test to compare experiments in GPT-2 mixed precision conversion\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 393, "deletions": 124}}, {"node": {"oid": "7d28b596f482f70052cc9e58374455bd5af7c92d", "messageHeadline": "Add function-body to opschema of FastGeluGrad (#9028)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 61, "deletions": 17}}, {"node": {"oid": "4322f7e647f5b4c751f7f42960fd450f0c7646d5", "messageHeadline": "Fix ROCm wheels CI pipeline break by installing latest protobuf from \u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 56, "deletions": 2}}, {"node": {"oid": "cf70635d2af3666f6630ac2d62ecfd6e63e48b0a", "messageHeadline": "Add Android executable drop in the Package pipeline (#9050)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 31, "deletions": 3}}, {"node": {"oid": "be8069869880c5981f4df3f8af5cc6e534e0ef11", "messageHeadline": "[js/web] a bugfix and add tests for wasm proxy worker (#9048)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 29, "deletions": 4}}, {"node": {"oid": "e574be4a5385226346e2d1e12704f6080e8af0e7", "messageHeadline": " [C API Docs] Add docs for run options tag/log level accessors/modifi\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 153, "deletions": 85}}, {"node": {"oid": "6036a6b915aacbd6c55c51db7dbb3ce81c47ce95", "messageHeadline": "Add type int64 for Equal, float types for ReduceSum (ROCm) (#9010)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 16, "deletions": 0}}, {"node": {"oid": "9174cbe3d5e40eb457a574c82e1968a50c0995a1", "messageHeadline": "Optimize CUDA Kernel for 3D and 4D Transpose (#8928)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 402, "deletions": 139}}, {"node": {"oid": "5969d576e551cf36f63e0497af7cb8b703541d02", "messageHeadline": "Revert \"disable half2 kernel by dfault (#9034)\" (#9044)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "34f37d29207108b9cba0215f899757f367733ed7", "messageHeadline": "Disable fallback for ortmodule api tests (#9018)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 17, "deletions": 3}}, {"node": {"oid": "c709380c5225deef0e637aae9735747c30b817c9", "messageHeadline": "Add full iOS job in package pipeline (#9036)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 79, "deletions": 30}}, {"node": {"oid": "1422a9ba6b2f09a65e9301873eda5a349cd1a8f3", "messageHeadline": "Remove previous temporary fixes and address TODOs (#9020)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 0, "deletions": 25}}, {"node": {"oid": "011cb8fd48b95bc952e50feef73137038d224cb0", "messageHeadline": "Fix Where op type reduction processing (#9033)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 19, "deletions": 7}}, {"node": {"oid": "a1021a1cf4bca68a975b2ebf4aed57f1b9a3fe4f", "messageHeadline": "Add BatchNorm kernel for ROCm (#9014)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 433, "deletions": 8}}, {"node": {"oid": "e83cc534d4bb8e7c76433adad010cce3cc6fc9d8", "messageHeadline": "Fix cmake POWER10 detection", "author": {"name": "Rajalakshmi Srinivasaraghavan", "email": "rajis@linux.ibm.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c674343d946c62e0045b7974331fe730f7146c8b", "messageHeadline": "Remove document text from error message in a couple of ops (#9003)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "c3321b17789e5e6c9f3ad18c298fe0b357f3a3c6", "messageHeadline": "Fix NVTX profiling so it can run in the shared CUDA provider (#9035)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 47, "deletions": 17}}, {"node": {"oid": "289999af351e3fea1568c2f8b97b9d3847f3ce82", "messageHeadline": "disable half2 kernel by dfault (#9034)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "8eb6546e8e94b417fbddd719377c177bf2fa645a", "messageHeadline": "enable eager mode with ortmodule (#8961)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 158, "deletions": 29}}, {"node": {"oid": "29d6573f3df4112c44c7c9a46db8cc291814d799", "messageHeadline": "Increase timeouts for Mac CI builds. (#9024)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b3c272586212c7771a57a0be4946af08d965e3f2", "messageHeadline": "fix cpuinfo compilation flag usage (#9029)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ce7b12bf5d0e73202a8d958bb0f235251ac9c4c8", "messageHeadline": "Added new fp16 allow/safe opcodes in PropagateCastOps (#8964)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 79, "deletions": 199}}, {"node": {"oid": "31af88c0bca5b5168e26082ab051d0b1112bedcc", "messageHeadline": "Update cross_entropy_loss symbolic for new argument from upstream tor\u2026", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "ff66cfdfa6e53203af11043db742f290bd7e89f7", "messageHeadline": "Enable linking in exception throwing support library when build onnxr\u2026", "author": {"name": "Zuwei Zhao", "email": "4123666+Zuwei-Zhao@users.noreply.github.com", "user": {"login": "Zuwei-Zhao"}}, "additions": 24, "deletions": 4}}, {"node": {"oid": "e5ee0b435db9007921adeadffe929f247a5d6055", "messageHeadline": "Attention Fusion for GPT-2 from Megatron (#8987)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 618, "deletions": 210}}, {"node": {"oid": "57b7ab56cd395a74d83d2a1523c9b32c42710f35", "messageHeadline": "Adding async fetching for webgl backend (#8951)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 145, "deletions": 14}}, {"node": {"oid": "5145fa236fdb88df9cab5368d18d9fb17a1c5354", "messageHeadline": "[js/web] fix ort web e2e test (#9025)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "2439ced3eca1f3b111427078a568893351613302", "messageHeadline": "API Documentation (#8948)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 4937, "deletions": 1159}}, {"node": {"oid": "6412c6a36209d4a939dced829e2f76b5179254e3", "messageHeadline": "do not add pkg wheel entry to the index html file if it already exist\u2026", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "e3570223620e95ffb4f658584e3932eae7aab406", "messageHeadline": "Remove onnxruntime team from CODEOWNERS (#8954)", "author": {"name": "Gary Miguel", "email": "garymiguel@microsoft.com", "user": {"login": "garymm"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "00fbc3b0bc45b0c2f88897c8f7c7280a4a54d028", "messageHeadline": "Instruct dockerfile users to do submodule updates", "author": {"name": "Spike Curtis", "email": "spike@lodestar.ai", "user": null}, "additions": 35, "deletions": 10}}, {"node": {"oid": "d78e90d1aff4123995609cae6f1c25a8542bc354", "messageHeadline": "Adding preprocessor checks for torch version during torch cpp extensi\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 17, "deletions": 0}}, {"node": {"oid": "0367e1f1c27ab097b4272f29e880843568c29e95", "messageHeadline": "Update Nuget Packge Pipline to CUDA11.4 and TensorRT8 on Windows (#9000)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "d209fe29b929f9620aaf1be1c3a67abb2e2aff61", "messageHeadline": "custom autograd func memory refinement  (#8993)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 100, "deletions": 34}}, {"node": {"oid": "d39959172f8b83b693e0e768b90c5fdab27fd47b", "messageHeadline": "Fix fuzz testing build blocking release. (#9008)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "1533f574e41ebda765e8544a630e8c3d63b414c5", "messageHeadline": "Add full Android job in package pipeline (#9009)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 158, "deletions": 74}}, {"node": {"oid": "c20cb766be3a756999127e1f684f8730b9521b8f", "messageHeadline": "Optimize sequence type usage on CUDA [3/n] (#9002)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 48, "deletions": 47}}, {"node": {"oid": "2e8792ca42413df5e9fb686387c6a1e7f755bc7d", "messageHeadline": "[js/web] fix karma launch with chrome headless (#8998)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ec63d10303d52f1bd9d1b56823da6241365f817b", "messageHeadline": "add model local function support (#8540)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 753, "deletions": 91}}, {"node": {"oid": "b7b42e0c5d4ae6f4052a89c3f3bc8ab03d9abe9a", "messageHeadline": "fast reduction for reducemean (#8976)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 130, "deletions": 41}}, {"node": {"oid": "1c872f9d7423a97e44274561167edb09f1866072", "messageHeadline": "Fix issues in TensorRT EP (#8996)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 27, "deletions": 5}}, {"node": {"oid": "6fbd0a823350615c613662f60cb27f18b4cbab24", "messageHeadline": "Change cmake_cuda_architectures to double quotes (#8990)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "5ae4c54ab8b8dfd2d204778fae4cf75e1d19483a", "messageHeadline": "Fix bug for validating GPU packages (#8997)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a30d9f5317c9bad0a82348bca978388bdbbf10e1", "messageHeadline": "fix windows gpu pipelines that use cuda 10.2 (training, reduced_ops a\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "450524359e6696f86a63ad950b0dbeb1c5cf86bb", "messageHeadline": "[js/web] WebAssembly profiling (#8932)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 108, "deletions": 11}}, {"node": {"oid": "0193490cbf2564d5da6a6f43d51a4b615616acab", "messageHeadline": "ReduceMin - add int64 cuda kernel support for opset12/13 (#8966)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 24, "deletions": 2}}, {"node": {"oid": "91c15843cdb8074e219510007cba0bb0e560a29b", "messageHeadline": "Fix a directml python packaging error (#8981)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e2194797a713f19a15ce2afbe0c6a78d5d8c467e", "messageHeadline": "bumping up to version 1.9 (#8982)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 24, "deletions": 16}}, {"node": {"oid": "00eca424132349641d9588a4bc6f4b06e8866ca4", "messageHeadline": "make_policy(SET CMP0104 OLD) (#8793)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "b7971575f82d0c91663a87551f1b8bbe9d010e50", "messageHeadline": "Fix python manylinux to not load cuda if it fails to load dependencie\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 85, "deletions": 82}}, {"node": {"oid": "0bb56a18cf183343fe2aa62dbcf3130fc3308c76", "messageHeadline": "Add TRT header file to ORT GPU nuget package (#8962)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 19, "deletions": 27}}, {"node": {"oid": "3be96f8a15e2b8b86fb448a6a7b9e669b173acaf", "messageHeadline": "fix: import error in TrtTable::Dict method (#8940)", "author": {"name": "senysenyseny16", "email": "82811840+senysenyseny16@users.noreply.github.com", "user": {"login": "senysenyseny16"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5d47b2e43134f15c7ca0b7a8adb30be055a47d65", "messageHeadline": "Add Einsum and Reciprocal op support in symbolic shape inference (#8931)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 167, "deletions": 38}}, {"node": {"oid": "60c98a86b7f0deb2bbb9f9ca8ab5d14e2fc383e7", "messageHeadline": "CMake file changes for macOS universal2 support (#8953)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 467, "deletions": 351}}, {"node": {"oid": "a9776d1c7050b514f014d902003167999d18a2ef", "messageHeadline": "Add QDQ model support in TensorRT EP (#8969)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 17, "deletions": 9}}, {"node": {"oid": "53eb79f9f6715d2af43dc68abcdde798565a1d69", "messageHeadline": "Gemm/Transpose fusion - additional pattern coverage (#8941)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 193, "deletions": 10}}, {"node": {"oid": "eebcc20f1052ef192ac4100923a840091b05dd91", "messageHeadline": "Add netstandard2.0 framework to nuget managed package. (#8960)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 25, "deletions": 20}}, {"node": {"oid": "a0c9408f0de5f071107e23fbf84c56285193a5d5", "messageHeadline": "Make TRT Version Configurable  (#8864)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 180, "deletions": 38}}, {"node": {"oid": "1f576e1766b0d16ba7ba4d23ecaf8a020329471b", "messageHeadline": "Detect necessary files inside GPU packages (#8955)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 336, "deletions": 188}}, {"node": {"oid": "a7f5bd226b7daaac8422e986a1b8d86a2b9b1602", "messageHeadline": "retarget torch181 to torch182 (#8947)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "0cc29095733cecc55efe0b8e0d8ff7cd2a9e427a", "messageHeadline": "Auto forward non method attribute lookups to the user's model and bin\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 375, "deletions": 8}}, {"node": {"oid": "c343f7cb43debe1d36aa745b79a18eed860dfa81", "messageHeadline": "Add Algorithm Search for ConvGrad (#8613)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 901, "deletions": 229}}, {"node": {"oid": "91f05f387a171c15a87ce6fe05e55eff7f40885c", "messageHeadline": "Update embed layer norm fusion to work with transformers v4.9 (#8914)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 611, "deletions": 253}}, {"node": {"oid": "e3489290193916c4e8db8867d441259d1e2cfb2e", "messageHeadline": "Minor cleanup from #7592 (#8952)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 32, "deletions": 29}}, {"node": {"oid": "5f30be3e924562dc087cc7f258bcf3999b89ad8b", "messageHeadline": "Exclude training support from BatchNorm in minimal build (#8939)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 37, "deletions": 16}}, {"node": {"oid": "47435311f43cf587b40da20dc33d94bbf950e1d4", "messageHeadline": "Include pytorch_export_contrib_ops in inference builds (#8878)", "author": {"name": "Gary Miguel", "email": "garymm@garymm.org", "user": {"login": "garymm"}}, "additions": 86, "deletions": 87}}, {"node": {"oid": "06bb2ec561ff1012fac50fafb7a90a125b40839d", "messageHeadline": "ignore direnv configs (#8861)", "author": {"name": "Gary Miguel", "email": "garymm@garymm.org", "user": {"login": "garymm"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "fe7f30aa142fff3163b4fc170afa206e1f5fba80", "messageHeadline": "Enable all-or-nothing fallback by default (#8911)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 144, "deletions": 33}}, {"node": {"oid": "1a34775fe97f16f332efa7d58ef5f44b716de066", "messageHeadline": "Fix the benchmark code (#8926)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 3}}, {"node": {"oid": "6490191f58640ce17d8026f210d773d8d3b96226", "messageHeadline": "Fix non deterministic of --input_int32 of transformer optimizer (#8927)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 89, "deletions": 39}}, {"node": {"oid": "7647caa520f4597abe2e3fd983c9ae8d612d0ae3", "messageHeadline": "update Tensorflow_Tf2onnx_Bert-Squad_OnnxRuntime_CPU.ipynb (#8898)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 327, "deletions": 277}}, {"node": {"oid": "4570d85f20fee18ea5d893b5b211142cf3a5ac2b", "messageHeadline": "Move setdlopenflags calls into _pybind_state.py (#8916)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 19, "deletions": 4}}, {"node": {"oid": "f711d8992aefefe7126b3aa7dd1522c944db594e", "messageHeadline": "Not to calc memory for inference (#8935)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "fbb6f0f5996ab97f3dff56abbb314df6b4e98204", "messageHeadline": "Fix an error in Nuget pipeline caused by merge conflict", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b058dee64896dde347558937dc0057cd0791d8fb", "messageHeadline": "Fix a couple of issues mentioned in the PR comments. (#8936)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "ddbc8bc5fc607adf0e79d28957a12ee9a46a9275", "messageHeadline": "Fix CPU Xor implementation (#8934)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "198561626291ea14865e7346bcdc30d98c6fdaa9", "messageHeadline": "Trim InferenceSession binary size. (#8917)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 189, "deletions": 87}}, {"node": {"oid": "332c2ba4f4fe0231eae3f33085bccc1237effa3f", "messageHeadline": "[js/web] Integrate ONNX Runtime Web CI with BrowserStack (#8859)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 185, "deletions": 0}}, {"node": {"oid": "757e9e6df78fc6a68a4d2f8c79fbb5c39aa22bdd", "messageHeadline": "do not post cuda version mismatch warning if cannot find local cudart\u2026", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f126a126997d94a2363461d05074b20478756b6f", "messageHeadline": "decouple pytorch from onnxruntime training build (#8815)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 19, "deletions": 29}}, {"node": {"oid": "9467f511ac12af99cfbf3c895d1e080f6d015192", "messageHeadline": "Disable some ORT graph optimizers in offline transformers optimizatio\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 21, "deletions": 5}}, {"node": {"oid": "225439193e484e724d96f89c19da2f3d8794e9c6", "messageHeadline": "Optimize Concat and Split on CUDA to eliminate host-to-device copies \u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 402, "deletions": 31}}, {"node": {"oid": "858989293dca5dbe7ce75ce11189e6b096add38e", "messageHeadline": "Reduce binary size of strided copy used by Concat (#8913)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 328, "deletions": 274}}, {"node": {"oid": "9e661b64ae59c976d72650faebec38bd3d0fdde9", "messageHeadline": "Fix cast propagation to not change casts from bool type. (#8925)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 42, "deletions": 5}}, {"node": {"oid": "6299a60bf8f021eb8137e29fac3a00c7b152494d", "messageHeadline": "Nuget: splitting PDB files to a separated package (#8903)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 139, "deletions": 105}}, {"node": {"oid": "00b0a9c1277922b81c28741e6c927692d8043289", "messageHeadline": "Add hugging-face models loss curve and performance guards to ROCm CI \u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 509, "deletions": 5}}, {"node": {"oid": "43d6951fa5d21e80adba6e5b3359412c9e407585", "messageHeadline": "Add warning message for combined trt +cuda python pkg (#8906)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "acd9db7fadb8da618c77aef31a4df626448343e9", "messageHeadline": "Fix location planning for initializers used only in nested subgraphs \u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 323, "deletions": 41}}, {"node": {"oid": "4dc0ddf606de526ef62afaeb8d285d4786ae3f1b", "messageHeadline": "support register external ep lib information (#8897)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 156, "deletions": 100}}, {"node": {"oid": "3eb08d4dc76f0c81a16a474b7e1e16e7a5e86067", "messageHeadline": "custom autograd func memory (#8901)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 181, "deletions": 7}}, {"node": {"oid": "feb747173e7a491c9b6942d14ead1cc1e22d015d", "messageHeadline": "[js/web] Update browser support table (#8900)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "8404a2d011749c277d89ddac324b99f6ed3ea8ba", "messageHeadline": "Add NNAPI E2E test for Android java package (#8912)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 34, "deletions": 8}}, {"node": {"oid": "a9a0d3f6fa822cd5a087a0fd20f7d517899a1e3d", "messageHeadline": "Update min supported macOS version to 10.14", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 9}}, {"node": {"oid": "70ca03d491e95b2799c1be6f3b4838290b086d4d", "messageHeadline": "Correctly set the skip check flags for ORTModule (#8891)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 60, "deletions": 36}}, {"node": {"oid": "69ab4670f753980aa751c0032d1ac457d2d57a45", "messageHeadline": "CUDA UpsampleNearest performance improvement (#7592)", "author": {"name": "Corentin Schreiber", "email": "54102755+cschreib-ibex@users.noreply.github.com", "user": {"login": "cschreib-ibex"}}, "additions": 40, "deletions": 21}}, {"node": {"oid": "129722db37292fe69eb21a1cd86ef7a4c6b2ec20", "messageHeadline": "Add android binary size monitor back (#8904)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 21}}, {"node": {"oid": "cd4b9f77536d113c381815c145ba39ad2275bea9", "messageHeadline": "Fix EP in transform (#8909)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "dc75a135c8fdcd7877e429fad9e87d5357c04772", "messageHeadline": "Add elementwise operators to DNNL execution provider (#8899)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 213, "deletions": 105}}, {"node": {"oid": "2e37fe3f68ebf91cde8fa70e3ad5bb31ccefbf9e", "messageHeadline": "Fuse HardSigmoid with conv. (#8674)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 195, "deletions": 21}}, {"node": {"oid": "206537936f78d6c592d1e4ac5c0777f4d4b459c5", "messageHeadline": "[js/web] enable proxy worker for wasm backend (#8862)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 771, "deletions": 335}}, {"node": {"oid": "33c0b3e94b101331ff59856c613943ff0b6675e2", "messageHeadline": "Perf test fixes (#8863)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 33, "deletions": 29}}, {"node": {"oid": "b7129305be21111ba9d503019e8b0ede9c6834e3", "messageHeadline": "[OpenVINO-EP] UEP v3.1 Release with OpenVINO 2021.4 (#8892)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 69, "deletions": 840}}, {"node": {"oid": "7659148d9f5d92aefc6ff0ec6e7db841a56fbd42", "messageHeadline": "Avoid round-trip copies for \"pass through\" subgraph inputs (#8702)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 385, "deletions": 31}}, {"node": {"oid": "42ba0c59318c9772f7aa529d35b10f9d88a52c78", "messageHeadline": "output related packages (#8886)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 8, "deletions": 34}}, {"node": {"oid": "f6e7cd805005f9dfdcf94dfdcaba659aac640c81", "messageHeadline": "remove existed dir for output external data (#8887)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "a171d13b197934450131e0173ed7a6454bd22251", "messageHeadline": "Add a formula for layernorm parity test (#8888)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 141, "deletions": 58}}, {"node": {"oid": "c6d9426ef2b36251cc03f220ec8a33f235706028", "messageHeadline": "Add binary size reporting back (#8883)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 71, "deletions": 102}}, {"node": {"oid": "868c8af9acfd1b841ca883315b73bc1c5c7f3c9c", "messageHeadline": "Abjindal/eager mode pipeline (#8870)", "author": {"name": "Abhishek Jindal", "email": "abjindal@microsoft.com", "user": {"login": "ajindal1"}}, "additions": 171, "deletions": 0}}, {"node": {"oid": "6df4e293ffbb47d739d2dedfbb87fa6234b8c37c", "messageHeadline": "Remove unused code in tools/ci_build/github/azure-pipelines/nuget/tem\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 76}}, {"node": {"oid": "7cd46cb9c4ee33b7fb5a7569727ede7456ad01d4", "messageHeadline": "Fix a problem in Zip-Nuget-Java Packaging Pipeline", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b75c1081ca46b626b9cc4a7b34b41db769c49ae9", "messageHeadline": "[Objective-C] Enable static analysis, second try (#8875)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 27, "deletions": 48}}, {"node": {"oid": "84f9271a8d5ff6e0a4de6005998250c680cfc6ef", "messageHeadline": "Enable registering external custom op schemas on Linux (#8889)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 84, "deletions": 46}}, {"node": {"oid": "03b680b9404c4722f3169465ecb13e253abded40", "messageHeadline": "Delete template.targets", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 22}}, {"node": {"oid": "fa27c19342ee18bba900c033619b86af017e0f4e", "messageHeadline": "Delete create_nuspect.py and template.nuspec", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 156}}, {"node": {"oid": "1b5909dea84651e20b5f50092829a4b4f73eac7a", "messageHeadline": "Delete download_cmake.py (#8885)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 90}}, {"node": {"oid": "c8dd0bf37ed73d3d7cd4a7135ce4dab4b7814fc0", "messageHeadline": "to publish stable wheel to ort channel (#8873)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 11, "deletions": 8}}, {"node": {"oid": "36fa0de8b729624d09094b22302781cd0e3ebc78", "messageHeadline": "fix regression and enable custom autograd func tests in CIs (#8868)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 30, "deletions": 80}}, {"node": {"oid": "6e20eb7eb328637672b34044905342b183356928", "messageHeadline": "Stop gradient for Multinomial, RandomNormalLike, RandomUniformLike an\u2026", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "df9438192a688d48a184a5d82689ee0fe5286e34", "messageHeadline": "Re-introduce saving of optimized onnx model (#8860)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 34, "deletions": 5}}, {"node": {"oid": "31926176ac44cf6733a7964276560c91e732e34f", "messageHeadline": "Support external custom operator schemas on Ubuntu (#8807)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 382, "deletions": 3}}, {"node": {"oid": "89e8bff121cb09b6f30f5b74ad85388fe29f7b21", "messageHeadline": "Enable selecting custom ops in onnxruntime-extensions. (#8826)", "author": {"name": "Zuwei Zhao", "email": "4123666+Zuwei-Zhao@users.noreply.github.com", "user": {"login": "Zuwei-Zhao"}}, "additions": 98, "deletions": 28}}, {"node": {"oid": "6ea9324f820974aa000dad7a945516a1b89b85b5", "messageHeadline": "fix EmbedLayerNormalization shape inference (#8876)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ae7f2d824df8183f5119c6d354f11b61a34de5b0", "messageHeadline": "Share the execution provider instance for training (#8719)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 734, "deletions": 436}}, {"node": {"oid": "6a1939252fad1ca2917b29a3530c7fdcd28ca368", "messageHeadline": "Fix Android java API failure (#8865)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 99, "deletions": 37}}, {"node": {"oid": "615df42b4627be5fb732018c4b93160dc436037e", "messageHeadline": "Add force_fp16_initializers in convert_float_to_float16 (#8871)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 116, "deletions": 198}}, {"node": {"oid": "0034ad72e6c6da0a343565fb3d5364a7cec53137", "messageHeadline": "Minimize changes to fix missing symbols used from C# (#8867)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 15347, "deletions": 186}}, {"node": {"oid": "f3083f4bf36cfc4a48bdff060a60ab7aa0f73a4b", "messageHeadline": "Support of sparse initializers with smaller indices data type (#8834)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 443, "deletions": 187}}, {"node": {"oid": "775f862067b305f8dcba3ff1d65d47c15119c30d", "messageHeadline": "Add new option to disable cpu sync for tensors  (#8490)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 22, "deletions": 5}}, {"node": {"oid": "6a477acecf912f3561349ce43eeb30bafdbebe53", "messageHeadline": "Add tensorrt_provider_factory.h to artifact (#8869)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "7e53a1df6f63953264e60134f49e06af15551940", "messageHeadline": "Enable selector action transformer infrastructure in minimal build. (\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 252, "deletions": 137}}, {"node": {"oid": "1886f1a737fb3aa891dea213e076a091002e083f", "messageHeadline": "Make SparseTensor infrastructure optional (#8802)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 593, "deletions": 67}}, {"node": {"oid": "cb59f46e042158a0b326e172e1ecf65af99d63fa", "messageHeadline": "Add gpt2 mixed precision conversion and parity tools (#8845)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 936, "deletions": 79}}, {"node": {"oid": "e8564d6597f562917825b82649b763f240ec5672", "messageHeadline": "[js/web] update emsdk to v2.0.26 (#8653)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 26, "deletions": 4219}}, {"node": {"oid": "a16c68110319e6d5b07e4c661b1546350345eabb", "messageHeadline": "[js/web] Prepare to integrate ONNX Runtime Web CI with BrowserStack (\u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 42, "deletions": 50}}, {"node": {"oid": "eb8f84e2a2dc2299e1c0c670a9c9e5065e7f7209", "messageHeadline": "Fix issue of GPU tarball/zip/java package (#8850)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 20, "deletions": 12}}, {"node": {"oid": "0cfc4ec09def573db8d5927a330bee3e4b1b067b", "messageHeadline": "[Objective-C] Enable static analysis (#8842)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 68, "deletions": 0}}, {"node": {"oid": "c325207f7a02f98033609302239cf0aa04cfacca", "messageHeadline": "Optimize MatmulGrad (#8846)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 8, "deletions": 29}}, {"node": {"oid": "ced2d8e59775aeadc0603d490b2845adbe4cb768", "messageHeadline": "Clean up TRT docker files (#8847)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 253}}, {"node": {"oid": "9cd7d836f7f4da77f29c1c770e56682ff3f05de8", "messageHeadline": "Delete Dockerfile.ubuntu_for_android (#8848)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 20, "deletions": 71}}, {"node": {"oid": "b21ea00020c7d9a1c3cfa3247ff9aeb2aacb579f", "messageHeadline": "Cleanup C# bindings to add EP (#8810)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 473, "deletions": 13930}}, {"node": {"oid": "613a6004715d6c80de6a5ee4e301edae9b4b894d", "messageHeadline": "relax android ci timeout to 180 minutes (#8844)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "32ecbf46918b36b8c30ef49d81e8dfd04c09c886", "messageHeadline": "Create combined GPU tarball and zip file package (#8827)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 253, "deletions": 53}}, {"node": {"oid": "cee79526fd68b6fb6b09d47dae0d48b47d2f7021", "messageHeadline": "Add opset 15 kernels for Pow, BatchNorm, and Shape (#8442)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 470, "deletions": 275}}, {"node": {"oid": "33a97e995bdea96acd39dc54903f764113ae1ac9", "messageHeadline": "POWER: Fix compilation issues with clang", "author": {"name": "Rajalakshmi Srinivasaraghavan", "email": "rajis@linux.ibm.com", "user": null}, "additions": 6, "deletions": 4}}, {"node": {"oid": "73fe7bfa0f59083c3bbe0853ad9b5c26a0f341f4", "messageHeadline": "Add ATenOp at::diagonal (#8838)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 77, "deletions": 0}}, {"node": {"oid": "237076a6605afb2a6ac4129a2e93ddc0a3c867c9", "messageHeadline": "Add option to disable FastGelu half2 cuda kernel (#8819)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 66, "deletions": 36}}, {"node": {"oid": "98ed235fc7e02387eb1596f7e8d9e8695f86fdd1", "messageHeadline": "Removed MSNPU code from eager. (#8832)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 0, "deletions": 16}}, {"node": {"oid": "4251e04eaef87800252c48389511b1628ead984b", "messageHeadline": "Removed assert (#8779)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "56b37e55e52b72148fab222cb3bdbe5be29c2dd1", "messageHeadline": "Add new transformers model type: Bart (#8698)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 282, "deletions": 21}}, {"node": {"oid": "3837027506134b9d37077d9793fd833c025ed5c3", "messageHeadline": "Remove pyopenssl from installation (#8830)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 15}}, {"node": {"oid": "ddd4586a2fbc047a1404ef78a7bd5063fc87f9ab", "messageHeadline": "[Symbolic Shape Infer] add more ops for auto merge (#8824)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7f1e8806492014d0c7e14cfbef791a67b7cd7549", "messageHeadline": "Reorder ORT eager headers (#8813)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "8992e31c8521492343aa22b683b0b25031d16e22", "messageHeadline": "Move iOS package from framework to xcframework (#8805)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 107, "deletions": 52}}, {"node": {"oid": "e25986781f9b3bfd363591ca2339374eb118d647", "messageHeadline": "Fallback to default quantization if quantization params is not found \u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 34, "deletions": 30}}, {"node": {"oid": "17b0664e34888c4d6396f88eeb1aa8b6400ef915", "messageHeadline": "Optimize sequence type usage on CUDA [2/n] (#8720)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 62, "deletions": 35}}, {"node": {"oid": "9053e1522dd23551411e2ff8f7d8c8084d1b6c4e", "messageHeadline": "Check for Python_EXECUTABLE in pyxir.cmake to fix Vitis AI EP build (\u2026", "author": {"name": "Jorn Tuyls", "email": "jtuyls@users.noreply.github.com", "user": {"login": "jtuyls"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "4bfff45859cdc09efc9920ea316f9e9fce76a72e", "messageHeadline": "Downgrade Eigen (#8817)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "2693af9799c6c26b2d1c29d722496330e4ca0940", "messageHeadline": "Ported changes / bug fixes from torch/ort. (#8784)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 280, "deletions": 163}}, {"node": {"oid": "f51f2bad66c529ba537cdef8b9ebc976ed437d42", "messageHeadline": "Fix for doxygen doc errors. (#8814)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 13, "deletions": 2}}, {"node": {"oid": "62c0d24340691900ef64bfefd336dbab3acfcac2", "messageHeadline": "Fix Windows Store build (#8753)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 166, "deletions": 96}}, {"node": {"oid": "ea68955c71c0e3c101caca1f6b467509ff6d6338", "messageHeadline": "Add more info to kernel registry manager hash lookup error message. (\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "d4a88cfe3f143d529df2a54f5eb3b4971ca933d7", "messageHeadline": "Add Gemm op to DNNL Exectution provider (#8799)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 331, "deletions": 321}}, {"node": {"oid": "89656bb7124f1329efcac3e807e881b89f0333a3", "messageHeadline": "[CoreML/NNAPI EPs] Move direct use of initializer data to unpacked te\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 189, "deletions": 101}}, {"node": {"oid": "0c5a30574237ed593cd297b95ff181d8f27be88a", "messageHeadline": "Bump up Nuphar cache version (#8806)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9fa0d8392a35e8fa0a7a94bd7e14d6c438c44e5b", "messageHeadline": "Extend node debugging utilities to push tensors and node placement to\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 463, "deletions": 37}}, {"node": {"oid": "4666a49106577723b9eb7e45f2eac7f84c0f632e", "messageHeadline": "Add Component Governance (#8794)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "19b82b438b6b7075df321284a79726b6208d344a", "messageHeadline": "GridSample OP implementation for CPU and CUDA (#8551)", "author": {"name": "XiyinOSS", "email": "60150203+XiyinOSS@users.noreply.github.com", "user": {"login": "XiyinOSS"}}, "additions": 735, "deletions": 5}}, {"node": {"oid": "6f2f4721ecac719c6329fc2491a4f3a7cdf43336", "messageHeadline": "Update Python setuptools classfiers to remove windows and mac (#8776)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 24, "deletions": 19}}, {"node": {"oid": "c117ac57b75735edfbbb57159371f6742405c9d1", "messageHeadline": "New S8S8 Neon kernel for arm64 only (#8783)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "94c3e2048bdeca11c83f85e0164da63f4187c884", "messageHeadline": "[convert_onnx_models_to_ort.py] Add option to specify NNAPI EP partit\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 22, "deletions": 5}}, {"node": {"oid": "81889a1cf6857efd81c55a496cccb6688bb1b9fd", "messageHeadline": "Invertible ReluGrad (#8773)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "2beb873c6bbf174e66f072d88ebd765bb94c6ff1", "messageHeadline": "move training CI agent pools to 1ES hosted (#8775)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 23, "deletions": 23}}, {"node": {"oid": "39059f2539f33c932513c8d37f4a1911005b66b1", "messageHeadline": "enable torch interop build (#8493)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 164, "deletions": 136}}, {"node": {"oid": "51152e1aaa98403ed250faaee4a6f21a292181a1", "messageHeadline": "Integrate TensorRT EP libs into existing GPU Nuget Package (Approach#\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 86, "deletions": 14}}, {"node": {"oid": "fe5046f48ef771110f972559be1c0264dcb5613f", "messageHeadline": "Add SparseToDenseMatMul to the list of required by test ops (#8774)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fa9fcb563466ec4bd7608a6bf4d3956602fa2ad2", "messageHeadline": "fixed the link (#8757)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b2813656f50d4c351528cf72fbadd9c423de8cbd", "messageHeadline": "eager: fix build against latest PyTorch master (#8745)", "author": {"name": "Aaron Bockover", "email": "abock@microsoft.com", "user": {"login": "abock"}}, "additions": 14, "deletions": 11}}, {"node": {"oid": "cb67fca738410abc629f517167593078d13d8908", "messageHeadline": "[js/web] enable 'use_ort_model_bytes_directly' by default (#8734)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 15, "deletions": 0}}, {"node": {"oid": "401de5911b6d37153ee1c3fb1efbc24fc7093c32", "messageHeadline": "Remove CUDNN dir from setup_env_cuda.bat (#8762)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b0c707caa871c820c7646138cd275a95c7db89c0", "messageHeadline": "[Nuphar] Do not handle MatMulInteger with zero-points (#8760)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 3, "deletions": 0}}, {"node": {"oid": "00b345eb7be34fa57d7b6addd012867d659fd452", "messageHeadline": "ARM Neon S8S8 kernel for QGemm (#8695)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1904, "deletions": 24}}, {"node": {"oid": "78759059f199ac109c63921b9a11704653174dd5", "messageHeadline": "[CoreML EP]Make coreml ep build on non-macOS platform (#8677)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 260, "deletions": 74}}, {"node": {"oid": "0983d619692a39a2135eddf5faa6d80a6c6bcb80", "messageHeadline": "refine glue code and tests (#8510)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 60, "deletions": 51}}, {"node": {"oid": "3406b7b5282468c73ca8df5970b4b5f720488e18", "messageHeadline": "Simplify UnpackInitializerData API (#8736)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 127, "deletions": 115}}, {"node": {"oid": "60089f7093e4e26f837be4bbf74d38cb97b43e4b", "messageHeadline": "Cuda11.4 (#8709)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 46, "deletions": 46}}, {"node": {"oid": "cc275e75294fa783ce3cb75aec5f16e18cb71821", "messageHeadline": "Gradient Accumulation optimization verified for correctness (#8273)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 300, "deletions": 55}}, {"node": {"oid": "224380448d371937a084b3d162be7964a702fa9a", "messageHeadline": "Expand Qgemm UDOT kernel to 8x8 block (#8562)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 1297, "deletions": 178}}, {"node": {"oid": "871eeb4dbd511a89f53c287a2f9f057153971cae", "messageHeadline": "Support dicts as inputs to ORTModule (#8718)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 244, "deletions": 32}}, {"node": {"oid": "ed254c283f1eba7dab8ed1cafc1bbf063bb45463", "messageHeadline": "Add support for experimental json config for fallback (#8759)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 85, "deletions": 22}}, {"node": {"oid": "6ecf626a9c491caffe3bd481d638b27d14534a6f", "messageHeadline": "[Nuphar] Parse node doc_string for quantize info (#8746)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 48, "deletions": 1}}, {"node": {"oid": "47b3ecb53b2fcde74cf6b7345b3407f6bc815f70", "messageHeadline": "Packaging pipeline now builds with PythonOp (aka running autograd.Fun\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 19, "deletions": 10}}, {"node": {"oid": "2b1f0816f835bd0fddc474c02239e16d51533a29", "messageHeadline": "to build cpu training packages for multiple multiple python versions \u2026", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "419834d2859038165986567568a3bcc3e8c81681", "messageHeadline": "Add PyTorch fallback for ORTModule forward exceptions (#8346)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 1596, "deletions": 459}}, {"node": {"oid": "11a618b2ecbe7c5fb627ad994462adff52a6fc7b", "messageHeadline": "Add engine encryption in TensorRT EP  (#8732)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 28, "deletions": 8}}, {"node": {"oid": "f668a795325073e5f47910b933bf537a73a7c970", "messageHeadline": "[js/web] fix perf mode in test (#8748)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "4ceedbe933f0e478e054397b84576dde1cdeb2e7", "messageHeadline": "[js/web] add SharedArrayBuffer check for wasm multi-thread (#8749)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "ae6fdd33332b4d0b45535b3c005e4486f1ec3601", "messageHeadline": "Bring code coverage dashboard back (#8394)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 63, "deletions": 139}}, {"node": {"oid": "0fb82f0f8ad7d3df36e8e73fdd644580818e479e", "messageHeadline": "Memory aware gradient builder.  (#8582)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 135, "deletions": 70}}, {"node": {"oid": "aa12d68c376ac63ee14b3dce62dbfd6c955c8f01", "messageHeadline": "Update ORTModule API docstrings (#8309)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 33, "deletions": 17}}, {"node": {"oid": "8713d76dd14baeb00328b9a20d246912bf70ea77", "messageHeadline": "Introduce C and C++ APIs for Sparse Tensors (#8621)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2335, "deletions": 155}}, {"node": {"oid": "8335d3dc0b13cc2d8d2c7539ddb9fc6265cdfc93", "messageHeadline": "Fix Python Packaging Pipeline (Training Torch 1.9.0 Cuda 11.4) (#8738)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9cefd1303bce7448c7c4adcb6898789918711f07", "messageHeadline": "Integrate Anubis (#8603)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 301, "deletions": 320}}, {"node": {"oid": "93e1e1dfa1fcca2022cb1b826b0986363dff456c", "messageHeadline": "Drop quant_util.h and move helper function into quantization.h (#8747)", "author": {"name": "Nick Kreeger", "email": "nick.kreeger@gmail.com", "user": {"login": "nkreeger"}}, "additions": 14, "deletions": 24}}, {"node": {"oid": "d0ff2621eeb2d965cfd8c5fc458d3f4615d6c16c", "messageHeadline": "[Nuphar] Fix Windows build in VS 2019 (#8728)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 6, "deletions": 8}}, {"node": {"oid": "8f7422be69eae4d9bd6f0a5535229624a926c5f5", "messageHeadline": "Limiting platforms where cpuinfo is included (#8716)", "author": {"name": "Chen Fu", "email": "1316708+chenfucn@users.noreply.github.com", "user": {"login": "chenfucn"}}, "additions": 42, "deletions": 35}}, {"node": {"oid": "e695cd304a11efdfb6c25ee5ef8d1094d89ef7b3", "messageHeadline": "Dnnl refactor (#8627)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 3829, "deletions": 6748}}, {"node": {"oid": "f04a235c77fe8e0171e351c6f8e355d50bcd5190", "messageHeadline": "Update manylinux build scripts (#8724)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 872, "deletions": 397}}, {"node": {"oid": "385b5718245a83a2693d668530b7905347e16024", "messageHeadline": "handle a corner case in removing useless reshape (#8669)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 23, "deletions": 1}}, {"node": {"oid": "d7baef765a5b52d12d92d6cdf76ef18310cf8ce7", "messageHeadline": "Fix 4byte to 8byte static warnings in attention_quant.cc (#8715)", "author": {"name": "Nick Kreeger", "email": "nick.kreeger@gmail.com", "user": {"login": "nkreeger"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "436ac6dd5f48c985be220ca864bda96cc9d59148", "messageHeadline": "Rename ml_value.h to ort_value.h (#8726)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 53, "deletions": 53}}, {"node": {"oid": "606b6271fa861652f9a86126523552bfed31daaf", "messageHeadline": "fix build (#8725)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "59e59b9e0ef45a8d96bb4ef972e3a0d745f8b4c4", "messageHeadline": "Fix unknown warning \"-Wformat-truncation\" build failure for arm  (#8721)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "26a7886a5dcbd5aca122c2f389cc4256f9d561ff", "messageHeadline": "Improve error reporting for posix system errors. (#8723)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 49, "deletions": 30}}, {"node": {"oid": "217b2c9f931b5b0f704df0c8336def47025d2148", "messageHeadline": "Removing filelock import from ORTModule (#8722)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "f00933c41a21b956dd4132c268c31f642d6853b1", "messageHeadline": "Update TensorRT parser to the latest (#8712)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "76d21bbeb205fc3c570b0ebf3de5ee66603a7614", "messageHeadline": "Update Android API level to 30. (#8717)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d9d0228d0bb60db9f36dc7b3561667a22264ade7", "messageHeadline": "[Symbolic shape infer] fix a bug in loop/scan (#8694)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "1a8adb96fe8a0d0879f0fb54c8dc692a7aa2f335", "messageHeadline": "Reduce templatization of C API and refactor for InitOrtValue (#8700)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 263, "deletions": 510}}, {"node": {"oid": "89601ee6b3e84464bfd3a1391bc68b5d6449bb79", "messageHeadline": "[EP Partitioning Utils] Add check for assigned node. (#8473)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 346, "deletions": 185}}, {"node": {"oid": "8d3c372dc9048d310199c14f01cf2262f467437f", "messageHeadline": "Fix typo", "author": {"name": "fredster33", "email": "64927044+fredster33@users.noreply.github.com", "user": {"login": "fredster33"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ce6675a74e415cf661d8a5d914bc3fbd3caf0f1b", "messageHeadline": "Avoid setting compile options on system libs for protobuf on Windows", "author": {"name": "Nathaniel McVicar", "email": "namcvica@microsoft.com", "user": {"login": "NathanielMcVicar"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5f74f198c1477039c23baff41f234f8f361d4ecf", "messageHeadline": "Merge CPU/GPU nuget pipeline (#8683)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 345, "deletions": 322}}, {"node": {"oid": "3e8cabbc3eb148ce706a284a6582ea13381f2570", "messageHeadline": "[js/web] WebGL backend refactor (#8586)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 5031, "deletions": 7408}}, {"node": {"oid": "7ff6a5e50374ff2d583f3788023cbe497ed14521", "messageHeadline": "work around build warning on jetson (#8701)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "333ef3c089906ac829630dfd0090c1db3727fb33", "messageHeadline": "Bump path-parse from 1.0.6 to 1.0.7 in /js/common", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "76dfe8108b439645d2e810bfc5aeb2ef16bdcaf8", "messageHeadline": "Optimize quantized LSTM (#8634)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 96, "deletions": 83}}, {"node": {"oid": "caacf249c57d51eb435156d286ad3d7cf3d01429", "messageHeadline": "Disable candy_opset9 WinML model test on Qualcomm Adreno (#8647)", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bec24ca4c17d7228f3a6c00df44d1a40aa54d905", "messageHeadline": "create packaging pipeline to support cuda11.4 (#8663)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 214, "deletions": 0}}, {"node": {"oid": "c6ef6b5bc8eeb49a6b8f6fadfd212c9e07ddde31", "messageHeadline": "Subgraph support for quantization tools (#8012)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 163, "deletions": 19}}, {"node": {"oid": "c5c5d3499b4b711d13d62959938f53915f51a707", "messageHeadline": "Rewrite dockerfiles/Dockerfile.arm32v7 (#8686)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 69, "deletions": 106}}, {"node": {"oid": "de2a53e46dcb0192c6961e3a91ffe7c3bbc8d34b", "messageHeadline": "[eager mode] fix build and support customize shared provider entry po\u2026", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 48, "deletions": 22}}, {"node": {"oid": "f661c1865413d5f8eda0dd45ed31f4829688b4fc", "messageHeadline": "Fix attention perf regression (#8682)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 69, "deletions": 47}}, {"node": {"oid": "c24335246b0838986ec842713a8ee4fc3db865a9", "messageHeadline": "Support bool type for Pad Op and fix Unsqueeze in Tile grad for Opset\u2026", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 15, "deletions": 5}}, {"node": {"oid": "a13daf550b88111434ec40efb67859b1decb2899", "messageHeadline": "iOS Coacopods spec fix (#8678)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "0725f80d2d22b31b1b1cc33444bc4a9808523dfe", "messageHeadline": "Revert \"Fix Windows Store build (#8481)\" (#8679)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 99, "deletions": 161}}, {"node": {"oid": "53e7831b5378e52f6a741e92dc733e50b5a3f912", "messageHeadline": "Fix Windows Store build (#8481)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 161, "deletions": 99}}, {"node": {"oid": "064a385b59598419c287a4b897afca376227d3bd", "messageHeadline": "Support int8 for operator Split (#8615)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 24, "deletions": 7}}, {"node": {"oid": "3a742f29108985e13c320598f5d32bee3edc655e", "messageHeadline": "Ensure that the BiasGelu op test runs on CPU. (#8671)", "author": {"name": "Nick Kreeger", "email": "nick.kreeger@gmail.com", "user": {"login": "nkreeger"}}, "additions": 7, "deletions": 9}}, {"node": {"oid": "e74f86059cbeb45b257b36506d27a0f60caf8bcc", "messageHeadline": "[js/web] enable SharedArrayBuffer feature for WebAssembly testing (#8\u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ed17ca3595e0b0fb3bf693e58195ab97543d8e09", "messageHeadline": "Remove onnxruntime/core/protobuf (#8617)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 3044}}, {"node": {"oid": "f0073308d0bf3ca6cbcf372b5907b4555f66e84b", "messageHeadline": "Fix CONCURRENCY_VISUALIZER build break (#8664)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "1a2b41dbbcec71214b1abed9ed77d11d57701aef", "messageHeadline": "packaging pipeline produces -cpu- named packages due to a logical err\u2026", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cadb43a715dfc4961bbd9b71e1aa8c3145f518ab", "messageHeadline": "Fix 'SyntaxWarning: \"is\" with a literal' issues in Python transformer\u2026", "author": {"name": "Mike Roberts", "email": "2947595+m-roberts@users.noreply.github.com", "user": {"login": "m-roberts"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "20f006c58053ee77d440ed4aa28837b07f85aedd", "messageHeadline": "Remove flake8 check from CMake build. (#8662)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 9, "deletions": 89}}, {"node": {"oid": "52a212e4f199238a6568185d3fa6eaf726aab7cc", "messageHeadline": "Bump ORT master version to 1.8.2 (#8646)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 18, "deletions": 12}}, {"node": {"oid": "3166a9b8e90162f32586d001a2e9b9a0748c32a8", "messageHeadline": "refine API of transformer optimizer (#8633)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 284, "deletions": 225}}, {"node": {"oid": "baf8c39a8d4d690ba0306ff555394a6a476e1e51", "messageHeadline": "Add Python checks pipeline (#7032)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 146, "deletions": 82}}, {"node": {"oid": "aaeb78113299cb34f4ece188a80f0fb6b4f13edd", "messageHeadline": "Revert \"Add netstandard2.0 target to OnnxRuntime.Managed (#8600)\" (#8\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6dd59a1117f714589bba74286ea1d7198edfa970", "messageHeadline": "revert onnx version (#8643)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a56e325eb823ac6ffc9406b7608159fa29d5686e", "messageHeadline": "constrain inputs for min/max grad UT (#8632)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "1ce4328846f0bc51f80201382c12656ec964e40a", "messageHeadline": "[js/node] fix js node install (#8650)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c86b813519789e39efb192d373554b4e56b4c3bc", "messageHeadline": "[js] resolve CodeQL warnings for force strict mode (#8645)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 55, "deletions": 0}}, {"node": {"oid": "419fd5cc6e8ab597fba76dbac6f54587a3bab95d", "messageHeadline": "reformat build suffix so that the latest is always correct (#8267)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 39, "deletions": 1}}, {"node": {"oid": "44ff80e816f4a2bbf74cb6b0de13ed4d30652b1a", "messageHeadline": "re-enable gpt2 fusion tests (#8566)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 116, "deletions": 80}}, {"node": {"oid": "1b902d0227d07f2fa69d40d1401cfeb3443a911a", "messageHeadline": "doc: add ort-web related instructions to update onnx doc (#8500)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 47, "deletions": 38}}, {"node": {"oid": "6d3c2c85ef0b663c9beead6db52e5c0551d639bf", "messageHeadline": "Integrate eager mode source code into onnxruntime repo (#8584)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 8227, "deletions": 1}}, {"node": {"oid": "484e9de55c109dadbeb552cd6ede21bbdd63b830", "messageHeadline": "Optimize sequence type usage on CUDA [1/n] (#8598)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 26, "deletions": 31}}, {"node": {"oid": "e791faeca553aee36c8524bf71444ed440aecb9c", "messageHeadline": "Fix bug in CPU force fallback logic  (#8597)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 34, "deletions": 3}}, {"node": {"oid": "f3a1aebb3385a446acf6239a1350e162dfaa1ccd", "messageHeadline": "[js/web] support override wasm file path (#8610)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 136, "deletions": 27}}, {"node": {"oid": "eab6c51413ac64eb904c0b7e1d54534df868cacc", "messageHeadline": "to create a training cpu package for torch-ort documentation (#7845)", "author": {"name": "liqun Fu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 100, "deletions": 0}}, {"node": {"oid": "045882194466abc0b9df1a54f7d2b2a151a19f41", "messageHeadline": "Delete linux-ort-srv-ci-pipeline.yml (#8628)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 28, "deletions": 33}}, {"node": {"oid": "24b14c650b3898c55cec8d1580cf4b5ae5688fe4", "messageHeadline": "Add parity test for LayerNormalization (#8622)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 437, "deletions": 231}}, {"node": {"oid": "dda9f53bed45e6e982adcee304011c383aff7196", "messageHeadline": "Build script logging updates (#8618)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "96eb9810ba5c3523e4fd640981be2d68c74355b0", "messageHeadline": "Update onnx (#8458)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 180, "deletions": 101}}, {"node": {"oid": "9d88b1de7836273dcf53d02d57947142d8f47384", "messageHeadline": "correct supported ONNX version (#8590)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e673d2e0504533541afde15c7da81d537dc5c871", "messageHeadline": "Add netstandard2.0 target to OnnxRuntime.Managed (#8600)", "author": {"name": "Georg Jung", "email": "github.com@gjung.com", "user": {"login": "georg-jung"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fa722d208b4af6f1c335273a1e465ae0f9d892dc", "messageHeadline": "[js/web] adding webgl pointwise conv kernel (#8418)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 226, "deletions": 38}}, {"node": {"oid": "1041fa34f49ec23cc06aea3635ad6cd9bbd409db", "messageHeadline": "Specify timeout for iOS packaging pipeline (#8616)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "f6da9ac6d639dda67e100ad3015c6f88b990ef40", "messageHeadline": "Add more detail to model IR version mismatch error message. (#8607)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "6c69baf78e8f301c2bbb234b868a21c6b3190c14", "messageHeadline": "Disable Training Windows GPU Debug build because it is failing (#8608)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1003db0058e1b08ebce89a25778816c75f610c47", "messageHeadline": "Fixing mlas unittest failures in POWER (#8498)", "author": {"name": "austinpagan", "email": "fossum@us.ibm.com", "user": {"login": "austinpagan"}}, "additions": 19, "deletions": 6}}, {"node": {"oid": "7b289a79275f59438310d63f86f5c4e7bd36baea", "messageHeadline": "Add test to evaluate Gelu and Fastgelu precision (#8592)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 292, "deletions": 5}}, {"node": {"oid": "375e86f0a0157904c2a0c51cc041e10258b9229c", "messageHeadline": "Make DNNL EP not depending on onnx (#8588)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 7}}, {"node": {"oid": "e09321f4dbec08f0e31a4e43879255a3a38826d2", "messageHeadline": "Update ORT format model conversion utility to optionally fail fast on\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "deab284e4cdd2ea57bd322ecf5ac7129b1c34e79", "messageHeadline": "fix build failure with --cmake_extra_defines onnxruntime_DEBUG_NODE_I\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "d14b08d09cee43b42fa9ac361a78d1cbdbceddef", "messageHeadline": "Update onnx-tensorrt parser and cgmanifest (#8585)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9e07ad93ae2a41ddd047f540b88f5b94cbea8cba", "messageHeadline": "[OpenVINO-EP 2021.4] Add/update Dockerfiles w.r.t OpenVINO 2021.4 Ver\u2026", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 165, "deletions": 91}}, {"node": {"oid": "717627775afc70d4d0fb1a31f4aaa39f177d67f2", "messageHeadline": "Increase build timeout (#8583)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ee99fb400c12b142c58233f6c15a310251320915", "messageHeadline": "Upgrade TensorRT to v8.0.1 (#8512)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 384, "deletions": 54}}, {"node": {"oid": "87975bdeef40ab79cb3f9d287e74af056cb8ef9c", "messageHeadline": "Use CUDA_HOME and CUDNN_HOME from the environment if they are not spe\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "49a6ff75e6ae34c31a9dfef98f3e43adde734d20", "messageHeadline": "Update py-packaging-stage.yml (#8569)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "d8c145d21853e6e304b01afe93823a0c309fdaf8", "messageHeadline": "[Nuphar] don't transpose B if A is a 1D array (#8568)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 28, "deletions": 8}}, {"node": {"oid": "0510688411db8cff48722fa9a8a3bd0d62dc0f45", "messageHeadline": "Update compliance tasks in python packaging pipeline and fix some com\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 527, "deletions": 421}}, {"node": {"oid": "330b8e74bdf551ea3bfcfa880c8e8bb3daf745e2", "messageHeadline": "Fix attention parity for GPT-2 (#8549)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 680, "deletions": 74}}, {"node": {"oid": "816ad86d14d0876cebcfac763795939d8f8eea3e", "messageHeadline": "Configuring ORTModule - Internal Options (#8537)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 445, "deletions": 3}}, {"node": {"oid": "c6f95841dceea4fa2f8f52991fa015e6e8781e30", "messageHeadline": "Add HardSigmoid to mobile packages. Used by PyTorch MobileNet v3 (#8552)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "464fd28ee9e30df317145d7b819413d1be350798", "messageHeadline": "Update iOS packaging script to default build static framework, disabl\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "ad093b94b9d95b1e1ef94c61eb60948cf65e1f7f", "messageHeadline": "Restore transformers tests and disable some tests (#8530)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 49, "deletions": 48}}, {"node": {"oid": "c80740b8fb959c61d32ae44ef0f66da64bf315ec", "messageHeadline": "Merged PR 6311990: RI 7/29 github master into dmldev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 13569, "deletions": 4454}}, {"node": {"oid": "d78568a9ac934d30f50a059e1fb4074da23ed77e", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 3223, "deletions": 475}}, {"node": {"oid": "0cf2ed029be11a4c63eae71b5548c4c2c87cb7cf", "messageHeadline": "Add python binding for CoreML EP (#8472)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 49, "deletions": 2}}, {"node": {"oid": "d243b3892962b70aa4da12ffea2f3bd3bdd137d8", "messageHeadline": "[Symbolic Shape Infer] Bump up required onnx ver", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 5}}, {"node": {"oid": "94c54718fbf2a6e0fbb5ce4cfcf43c5a056be2b7", "messageHeadline": "fix build break (#8536)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "778680202b8836db96b49678e6ac9a3f28e003c5", "messageHeadline": "remove unused functions to avoid warnings", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 0, "deletions": 54}}, {"node": {"oid": "5e2f4263db73046d79815a81ea80fec6a53df537", "messageHeadline": "Enable cast propagation in the frontend. (#8517)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "00d8f8ce95e3ea8b7ad45b3e96bf615a8e94f66e", "messageHeadline": "enable shared lib based execution provider test on linux (#8480)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 19, "deletions": 10}}, {"node": {"oid": "2e28cbaa64e06fc32f86aa7481526139fdaf9c33", "messageHeadline": "Configuring ORTModule - End User Facing Options (#8470)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 337, "deletions": 88}}, {"node": {"oid": "6f5bf8b8f2a76dcfca3d2d45289e20227f043930", "messageHeadline": "Update Linux Training CPU CI pipeline (#8518)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 60, "deletions": 5}}, {"node": {"oid": "1370cbe256842ee8d0b1f63f56b8b87bf48f507f", "messageHeadline": "[ORTModule] Extract output schema in module's true train/eval mode (#\u2026", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 52, "deletions": 10}}, {"node": {"oid": "a71dab691d1d67469654e8e35043aa8bceea8b9e", "messageHeadline": "Implement BatchNormInternal for cuda (#8172)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 585, "deletions": 86}}, {"node": {"oid": "539d1d44c1eea238c21296f81dc18c2d4485eb95", "messageHeadline": "Optimize ARM64EC build (#8515)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1154, "deletions": 37}}, {"node": {"oid": "17986985453dec2251985be8c72e0d468cc9d37c", "messageHeadline": "avgpool2d atenop (#8507)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 43, "deletions": 8}}, {"node": {"oid": "73660d78df4972e393698dc6f7b1f71b8ee57c4e", "messageHeadline": "Fix WinML build warnings in HStringFromUTF8 (#8519)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "ceeb1a65d64c00e5ce3b12e1c1c53f4308072cd9", "messageHeadline": "Add quantization support of GEMM directly with QGemm (#8447)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 743, "deletions": 68}}, {"node": {"oid": "0f46b0864626349a4582ae65cd57e25eaef85977", "messageHeadline": "improve the qlinear avg pool perf (#8514)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 114, "deletions": 25}}, {"node": {"oid": "56441dcd881ec0c71a90b48e4c4a88a179160927", "messageHeadline": "Limit work items to available threads, upgrade checks from assert to \u2026", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "686f9b530b14bfab4eea5488533d92fa6725b33b", "messageHeadline": "ORTModule set_seed in int (#8511)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "7d47175f76a0dc71268ba41f2f4c37a05fbda4a8", "messageHeadline": "cleanup NCHWc transformer (#8479)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 49, "deletions": 63}}, {"node": {"oid": "3850755feb41d19ac26d9079c6f539220810d6c9", "messageHeadline": "Fix: onnxruntime_eager library does not compile on Windows due to pat\u2026", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6b05a625842dd8e23b01753c33c50c8c44666542", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 10348, "deletions": 3981}}, {"node": {"oid": "1685ab81381fdc4d6948bcd19c5729c06a0cd49c", "messageHeadline": "Implement Concat with Strided copy  (#8336)", "author": {"name": "Oliver Rausch", "email": "Oliverrausch99@gmail.com", "user": {"login": "orausch"}}, "additions": 807, "deletions": 39}}, {"node": {"oid": "4c939e1cb7b24fab71602e0cd63fcf62bbcb5b92", "messageHeadline": "Add an option to use the input model bytes (ORT format only) directly\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 90, "deletions": 29}}, {"node": {"oid": "1ae32655b3c642dee594bfc7fdd7bd3ddfc89c0c", "messageHeadline": "fix t5 assert error (#8501)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 49, "deletions": 0}}, {"node": {"oid": "b4baac888c7fd9b338ad7af875422f0cf3a88a30", "messageHeadline": "[NNAPI EP] Make partitioning stop ops configurable from Python API. (\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 49, "deletions": 43}}, {"node": {"oid": "421c4059c0b7ea4fb6347b7f7b1ba4228b6e18cc", "messageHeadline": "[iOS Packaging] Update build definition (#8503)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "0a70c2de0074ec6b48234e5a462e312e82be9aec", "messageHeadline": "[Nuphar] Add support for opset 14 (#8483)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 457, "deletions": 232}}, {"node": {"oid": "91936864ce823b16207e9c3dad5958c8adcea73f", "messageHeadline": "Expose additional shared_provider APIs (#8478)", "author": {"name": "Ankur Verma", "email": "31362771+ankurverma85@users.noreply.github.com", "user": {"login": "ankurvdev"}}, "additions": 13, "deletions": 9}}, {"node": {"oid": "534c22d7694d625158f0ebc7a9929529239ca8e9", "messageHeadline": "use float for alpha in attention Gemm (#8477)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 53, "deletions": 5}}, {"node": {"oid": "a9fc3c448c0903dd084a8f083493af7672f7e25b", "messageHeadline": "Improves documentation, show InferenceSession contructor attributes (\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 41, "deletions": 8}}, {"node": {"oid": "79097ef5535cc5ac18fc8e9010c99de08df21340", "messageHeadline": "remove useless reshape node (#8419)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 25, "deletions": 0}}, {"node": {"oid": "6dee9b9d2d2be596276d7fcc932362a52f7d059f", "messageHeadline": "attention fusion kernel refactoring (#8432)", "author": {"name": "Viswanath Boga", "email": "44417868+viboga@users.noreply.github.com", "user": {"login": "viboga"}}, "additions": 50, "deletions": 72}}, {"node": {"oid": "a396c9e572eaa6a652e58d7609acc379f4b7eda4", "messageHeadline": "Add more safety checks to the C API (#8474)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 16, "deletions": 2}}, {"node": {"oid": "6a07172a9309f79c37f11e71ea1829718d17d160", "messageHeadline": "Restore cpu affinity after loading tensorflow model from transformers\u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 43, "deletions": 0}}, {"node": {"oid": "e66846da4a5afc9337a4cd83336075ce1f866c65", "messageHeadline": "revise terms according to guideline", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 28, "deletions": 28}}, {"node": {"oid": "ab5289f10972b9910b15abdb291b232852a473a2", "messageHeadline": "Performance: enable faster training with skip checks config (#8411)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 73, "deletions": 33}}, {"node": {"oid": "c8d210de296ba3c549e99725d9d454ab8ab23be0", "messageHeadline": "Decouple Forward and Backward of ATenOp (#8301)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 726, "deletions": 1417}}, {"node": {"oid": "619a8782a508563dd180a17d373d77875d4fa18e", "messageHeadline": "Improve AddValueInfo (#8451)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 12, "deletions": 15}}, {"node": {"oid": "b2b9de939f92802022caaea0e54ffd1c2e6e3282", "messageHeadline": "cleanup onnxruntime_mlas.cmake of old gcc workarounds (#8469)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 18, "deletions": 90}}, {"node": {"oid": "002e427c5bd0d60d765775382b6fd24bc4a76db5", "messageHeadline": "Add UINT8 datatype support to Java (#8401)", "author": {"name": "Frank Liu", "email": "frankfliu2000@gmail.com", "user": {"login": "frankfliu"}}, "additions": 46, "deletions": 1}}, {"node": {"oid": "950fe5e28b0b0fafb2fc68f2b0910ece0650ae31", "messageHeadline": "Implement SparseTensor and infrastructure suppport and advance ONNX c\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 5651, "deletions": 587}}, {"node": {"oid": "f641c0f4e8e2076550aeaba6d936e351b158d460", "messageHeadline": "Update requirements.txt", "author": {"name": "raviskolli", "email": "48601275+raviskolli@users.noreply.github.com", "user": {"login": "raviskolli"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9073c094d40730b91470ef16869cadb6f42f825b", "messageHeadline": "Update torch litghning and re-enable test", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "e8ee31bcc3584d2f7758ba013659fa273841acc5", "messageHeadline": "Update onnx_model_bert_tf.py (#8457)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "9a6fa057c8ee536a75c13eba2968e11718accbe9", "messageHeadline": "[Java] Allow extraction of multidimensional String tensors (#8452)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 95, "deletions": 17}}, {"node": {"oid": "287a2a778f14abb4ba47ffe384a5554c2725e018", "messageHeadline": "Update CODEOWNERS with mobile team ownership of expected kernel def h\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "3360024a0b2a55a9093ef477b097e4bbaa612cb5", "messageHeadline": "Support plugging in custom user-defined allocators for sharing betwee\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 648, "deletions": 420}}, {"node": {"oid": "989491c333a31fa6bf2b3d1e9e430d5c6552cf7c", "messageHeadline": "[NNAPI EP] Make partitioning stop ops configurable. (#8444)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 160, "deletions": 13}}, {"node": {"oid": "892ac9f55add28d46845f5acca8b10394c6b3748", "messageHeadline": "code structure update (rename only) (#8410)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 24, "deletions": 31}}, {"node": {"oid": "4275055868359e41f436c2b170f03d2a52fda057", "messageHeadline": "Add Gridsampler contrib op (#8372)", "author": {"name": "DeyuHuang", "email": "10047193+hwangdeyu@users.noreply.github.com", "user": {"login": "hwangdeyu"}}, "additions": 278, "deletions": 1}}, {"node": {"oid": "53d5814d12da3c684a9352c45bf42c68e5fb901d", "messageHeadline": "Move the wrapped types out of provider_interfaces (#8455)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 903, "deletions": 900}}, {"node": {"oid": "14b045ad5260d5b79d18804af82261f09e2296b9", "messageHeadline": "Add link to sample repos (#8417)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 12, "deletions": 19}}, {"node": {"oid": "83bb771e0b0aa3035b793bf489d881ce80e06212", "messageHeadline": "Merged PR 6283367: RI Onnxruntime github to internal Fork 7/21", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 26379, "deletions": 23313}}, {"node": {"oid": "b0c0b087a411ec10f328a1fcfdc7c5010d53beb1", "messageHeadline": "FIx merge conflict", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 26379, "deletions": 23313}}, {"node": {"oid": "695536a7ac5d97cac911d7cba324b883ac974756", "messageHeadline": "Make some common macros safer to use. (#8445)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 19, "deletions": 14}}, {"node": {"oid": "972aee83080b9d1bb64ed1d8f0ef8b2266f32f1a", "messageHeadline": "Fix GCC build error in quantization tests (#8449)", "author": {"name": "Oliver Rausch", "email": "Oliverrausch99@gmail.com", "user": {"login": "orausch"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "7e2ecb2eebdca78353b3c6aad24b0eb1ba92df8c", "messageHeadline": "Remove unnecessary line as no headers exist now (#8446)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "55b26b69513d3972e7c0a70f0d54f3fbb8adc054", "messageHeadline": "[Java] Adds support for DNNL, OpenVINO, TensorRT shared providers and\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 341, "deletions": 82}}, {"node": {"oid": "1cd9b47d8d9c56a3fce40b9166c64bde6920c780", "messageHeadline": "Remove all C/C++ samples from our C# dir (#8441)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 817}}, {"node": {"oid": "894fc828587c919d815918c4da6cde314e5d54ed", "messageHeadline": "POWER10: Additional check in cmake", "author": {"name": "Rajalakshmi Srinivasaraghavan", "email": "rajis@linux.ibm.com", "user": null}, "additions": 13, "deletions": 1}}, {"node": {"oid": "28527b48679cc6990382c206e28f978f92887824", "messageHeadline": "Handle duplicated names for output_grads (#8431)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 70, "deletions": 0}}, {"node": {"oid": "cc9f793b486fee35919877ef7e02c6f96070e1a1", "messageHeadline": "Move one function from cuda_provider_factory.h (#8407)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 13, "deletions": 50}}, {"node": {"oid": "3e7fcd8c92eec8bc8cb8b867403bc1c9a9630d35", "messageHeadline": "Fix iOS packaging pipeline failure (#8433)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "bf54fe481e5b57a178329737ce41b0121591cb21", "messageHeadline": "[CoreML EP] Support 1D Conv for coreml ep (#8398)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 78, "deletions": 9}}, {"node": {"oid": "862bc8c7a032ecb7fe6098d5feb56981ff17bd6f", "messageHeadline": "shape infer for present output of Attention op (#8430)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "0f989c6162583774a5905f15a973c389123c259c", "messageHeadline": "bumping onnxruntime version to 1.8.1 (#8429)", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 21, "deletions": 15}}, {"node": {"oid": "e04e1d5ce0b6e1dbcc9924515c413b50ff368d0b", "messageHeadline": "Move shared providers CPU providers into separate file (#8293)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 495, "deletions": 452}}, {"node": {"oid": "dfe42e185c6c6de68177db8ecf307645ce831aec", "messageHeadline": "update bert notebook to use onnxruntime 1.8.1 (#8379)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 675, "deletions": 804}}, {"node": {"oid": "afce0e2543f04f024e761946dd9abbde93bc6d89", "messageHeadline": "Attention kernel update to handle different Q,K,V hidden sizes (#8039)", "author": {"name": "Viswanath Boga", "email": "44417868+viboga@users.noreply.github.com", "user": {"login": "viboga"}}, "additions": 572, "deletions": 119}}, {"node": {"oid": "c3129306e582cb0a3e77d54f44ec5c04939c1494", "messageHeadline": "Enable string attributes for experimental model building (#8428)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "1686e8ff57d86a8728033a244754fd7b7714edad", "messageHeadline": "[OpenVINO-EP] 2021.4 Release (#8369)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 317, "deletions": 66}}, {"node": {"oid": "8544c222cecd41a069a3356d4cadc7dd18cb6196", "messageHeadline": "Fix build error about `if` being constant in Eigen and Protobuf (#8291)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4a614637a79e41826236e9528b2a374c49a66341", "messageHeadline": "[NNAPI EP] Update NnapiExecutionProvider::GetCapability() to use part\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 387, "deletions": 431}}, {"node": {"oid": "2f408f757ec54fb1659bca74f358795547030c33", "messageHeadline": "Improve performance of Pad CUDA kernel (#8408)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 205, "deletions": 10}}, {"node": {"oid": "127b1f0d01c4e255e5f1f076fd487ddca4fad220", "messageHeadline": "SDL Native Warning errors in DML ep (#8396)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 1273, "deletions": 656}}, {"node": {"oid": "bcd50afafbb95d9206535802669a7b4a9ec5a019", "messageHeadline": "[CoreML EP] Fix failure for layer without name (#8399)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 42, "deletions": 23}}, {"node": {"oid": "963d883de8dfdcf658a2c79a115f4542480f848f", "messageHeadline": "Create a common directory for quantization code and functionality. (#\u2026", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 825, "deletions": 385}}, {"node": {"oid": "ef930b3ca980b78766d76f57139a7a9b00a377b2", "messageHeadline": "[Objective-C API] Fix ORTIsCoreMLExecutionProviderAvailable link erro\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "c5038063ed0e72423c00097c499a67afac773b0c", "messageHeadline": "Add iOS/macOS static framework (#8357)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 208, "deletions": 199}}, {"node": {"oid": "41f1280fc91adb7676891e377b86b20ece7a54b1", "messageHeadline": "Fix transformer optimizer (#8392)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 50, "deletions": 37}}, {"node": {"oid": "88d1ffe9b805ba2921ec8ce92a20846da687acce", "messageHeadline": "Fix invalid access in log call. (#8389)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "0a1c00e8dbd908cb4418462d3d1e168ea800aaa5", "messageHeadline": "[js/node] remove unused dependency node-pre-gyp-github (#8388)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 269}}, {"node": {"oid": "5cd254aa79662136876798878b7476043505b42f", "messageHeadline": "update gpt2 attention fusion for past pattern (#8375)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 531, "deletions": 104}}, {"node": {"oid": "4e1c5f6ef4cc5db61184437164fae8f66f175546", "messageHeadline": "Move the samples to a new repo (#8374)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 8390}}, {"node": {"oid": "4931ef666d5a2c47bd1a8b1b23743d72f29c217b", "messageHeadline": "Update ORTModule frontend code owner file (#8335)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "68c5eb541482d65d8bc4d6680188050bde1a4b18", "messageHeadline": "Fix reduced ops CI failure (#8377)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "e340a599930833919ceb76a1288d5edd51ac10d5", "messageHeadline": "Update machine info script for transformers notebooks (#8376)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 40, "deletions": 13}}, {"node": {"oid": "16f69042329dcc1458d52f7174b52ff9913bb36a", "messageHeadline": "[iOS] Packaging pipeline improvements. (#8324)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 36, "deletions": 21}}, {"node": {"oid": "0020703d001267a1cb2b48f666fe221daf223ee6", "messageHeadline": "Fix cpuinfo initialization failure in mlas test (#8366)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 40, "deletions": 51}}, {"node": {"oid": "04297110c3e0eb317b6bfd2e4476da7d3b0fe362", "messageHeadline": "Support int64 in ReduceMin cuda op for Opset 14 (#8307)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 78, "deletions": 29}}, {"node": {"oid": "8d8db7c9f0e73f9ce2e4702212cf403ef7c2c852", "messageHeadline": "[ROCm] clear last status if hipErrorNotReady (#8358)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 39, "deletions": 12}}, {"node": {"oid": "178c1397186e181016a883cc948fb22d34555a49", "messageHeadline": "cleanup formatting in skip_layer_norm.cc (#8371)", "author": {"name": "Nick Kreeger", "email": "nick.kreeger@gmail.com", "user": {"login": "nkreeger"}}, "additions": 32, "deletions": 30}}, {"node": {"oid": "31f291f0afa77aa5b96d8082bab57dadc8c11e51", "messageHeadline": "Add TRT EP memory leak test into trt perf script (#8155)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 511, "deletions": 1}}, {"node": {"oid": "eda1411e034731fb35ecc8c6d9749e573dbfc1be", "messageHeadline": "Fix symbolic shape inference regression in RoBERTa training (#8364)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 27, "deletions": 0}}, {"node": {"oid": "7db4fc8c2a824aa7bac82306abc69e7fd7a10813", "messageHeadline": "Fix segment fault for custom function (#8331)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 40, "deletions": 9}}, {"node": {"oid": "5bf862eef946a737420ad5c5203f3f14fc5760b0", "messageHeadline": "Fix build break on windows arm64 (#8361)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "530d7bb46db40bbb2932e3449f732c569d896595", "messageHeadline": "Temporarily disable transformers tool test (#8360)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "0a5b75f5cdbbddc69fed00b2ecc0f27dcaf60747", "messageHeadline": "Update submodule onnxruntime-extensions. (#8282)", "author": {"name": "Zuwei Zhao", "email": "4123666+Zuwei-Zhao@users.noreply.github.com", "user": {"login": "Zuwei-Zhao"}}, "additions": 104, "deletions": 1}}, {"node": {"oid": "29ca08a729ede57af629ba94d2783aed40566d9a", "messageHeadline": "Update Dockerfile.cuda: remove compute capability 30", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "df4cb6f30175b29ccf563a17dfd36cfa3670b186", "messageHeadline": "Adding pytorch cpuinfo as dependency (#8178)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 215, "deletions": 10}}, {"node": {"oid": "eec8e1394a2e633364be2d53f36587ed99743e79", "messageHeadline": "Memory map files on windows to speed up model load (#8349)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 48, "deletions": 20}}, {"node": {"oid": "f6956e0259567a6fdd325014795b8712f9df8ad2", "messageHeadline": "Refactor qgemm file (#8322)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 3204, "deletions": 3054}}, {"node": {"oid": "b7c9696ac30b25c4fc1fcc7858719b97b00d747c", "messageHeadline": "Symbolic_shape_infer fixes (#8280)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 226, "deletions": 46}}, {"node": {"oid": "10142f95107a15d5f8228a9ae9fe4c58db82ed7a", "messageHeadline": "Add metadata_props to ORT model (#8340)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 284, "deletions": 13}}, {"node": {"oid": "60641a19e40c5f4c90395eb6d5a4675588f545a5", "messageHeadline": "Add \"/external:templates-\" to VC++ flags (#8338)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e467d78a117a6c7e5d3fb5c09a9ebc949f52a2de", "messageHeadline": "fix a typo (#8334)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "598454bb5fceff02800997262a0599522fb9c2d8", "messageHeadline": "Fix the mix precision handle for square case (#8333)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 14, "deletions": 13}}, {"node": {"oid": "187743726b29e1c994c965ff2bde4bc36e8f98d2", "messageHeadline": "[CoreML EP] Add Int32<->Int64 handling around coreml ep (#8183)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 101, "deletions": 43}}, {"node": {"oid": "5369821ad67ef41ec0be733417203eaf046592c8", "messageHeadline": "Support SpaceDepth ops in the CUDA and ROCM EPs (#7960)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 465, "deletions": 79}}, {"node": {"oid": "1b2e1a7e0c7b706850b7835db55173907a24ae25", "messageHeadline": "Refactor QDQ optimizers to enable future usage in minimal build (#8191)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1756, "deletions": 756}}, {"node": {"oid": "46e5c8d4b9eb01ecc5a35fa389d9fdc4ba30075f", "messageHeadline": "Cosmetic change in test infrastructure (#8292)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 135, "deletions": 118}}, {"node": {"oid": "5454af4b95c9a8ad08e8f03633e03b14447d93f3", "messageHeadline": "decouple the shared python dependency  (#8294)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 60, "deletions": 107}}, {"node": {"oid": "067759b387fd1f789d0dc4755b9d7fd3f7063bf9", "messageHeadline": "Fix bad URL to huggingface onnx-export example notebook", "author": {"name": "Dmitry Yutkin", "email": "d.yutkin@corp.vk.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "84bc20fe9d09542a6ec9dd9b97ac960161fcecfd", "messageHeadline": "Enable cast propagation with level one by default. (#8286)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f40df3021949962ef8cc66fe1de97f752d552677", "messageHeadline": "Replace functions with secured version for OSX compliance (#7586)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 18, "deletions": 10}}, {"node": {"oid": "6dbfb8db0e0872f83832660f84a84598c4e2d96c", "messageHeadline": "autograd function fallback perf (#8312)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 212, "deletions": 20}}, {"node": {"oid": "c254c3c355c02f12e9b3728aeedf5a66d48cdf4c", "messageHeadline": "Fix issue with ONNX to ORT format model conversion script when given \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "6652d17dcd252abd3bbd2d3077b46433617f202e", "messageHeadline": "Support lists as inputs to ORTModule (#8311)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 113, "deletions": 1}}, {"node": {"oid": "9a855fe9e728be608281281df61fc0c9cd687795", "messageHeadline": "Make Torch CPP extension build optional for packaging pipelines (#8305)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 18, "deletions": 18}}, {"node": {"oid": "d7c3703371f8d4c79733c89644381bb673dbf171", "messageHeadline": "handle unsqueeze change in opset13 (#8308)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 37, "deletions": 6}}, {"node": {"oid": "2347a0aca85f81ad8b30b2eb9f41f9b01bc479c2", "messageHeadline": "Autograd Function Fallback bug fix - moe support (#8105)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 273, "deletions": 89}}, {"node": {"oid": "40e5279f8faf18be0281a66709f328b93054669e", "messageHeadline": "Drop unused functions from math.h (#8304)", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 7, "deletions": 35}}, {"node": {"oid": "62d1458ea8effcf1d82590c0ea738c1ed488bb04", "messageHeadline": "Move kernel implementations outside of lookup table utility functions\u2026", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 177, "deletions": 142}}, {"node": {"oid": "090bae21ab9fad330e1b5572f9e9d2bebca69d7b", "messageHeadline": "Pinning pillow version to 8.2.0 to circumvent regression introduced b\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "008c5f7640eb53c31fa063b1348cd5708746239e", "messageHeadline": "Use single builder image across Python versions for ROCm wheels (#8302)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 61, "deletions": 44}}, {"node": {"oid": "56e4dd1d3e27a1ac16afe16ac02261cdde40057d", "messageHeadline": "Fix optimizer crash (#8274)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 147, "deletions": 7}}, {"node": {"oid": "e71846b029bb2c32248936fc74fdc7c323863b25", "messageHeadline": "fix ld_preload for rocm (#8290)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 2, "deletions": 1}}, {"node": {"oid": "036eee5b662497c061f8588517befabbcbf9d5b0", "messageHeadline": "register softmaxinternal with rocm (#8289)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 8, "deletions": 0}}, {"node": {"oid": "969eb545d1a98ed867d35d4f77b0cd6a0fbfcbc1", "messageHeadline": "Update issue template to ask users to check known issues to avoid rep\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0fa9ac36484b3fc123225e220aef315110af4603", "messageHeadline": "Remove path from telemetry strings (#8281)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 14, "deletions": 9}}, {"node": {"oid": "552806f3be517620c2ab006f36c4373b79386545", "messageHeadline": "Fix lamda function formatting in layer_norm.cc (#8276)", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 37, "deletions": 35}}, {"node": {"oid": "2bda2a62fddea0eb60369acf5d141c958dde084e", "messageHeadline": "Pin version of Pillow to 8.2.0 to circumvent noncompatibility with nu\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "88ec95ea967c7f148a526fd33ee1bf0c365763d7", "messageHeadline": "Support OrtMemTypeCPUInput for ATenOp/ATenOpGrad (#8116)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 311, "deletions": 159}}, {"node": {"oid": "b42e7d2c785e8825223f818b3884d41f9c4b4073", "messageHeadline": "Add iOS packaging pipeline (#8264)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 332, "deletions": 69}}, {"node": {"oid": "a9a2394fa5b7ab0f8b55b2ecfeefb91967eedff9", "messageHeadline": "disable computation reduction optimization for non-gpu build (#8251)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "9cfe642b342f5405f9da370ca09292ff381ea7c8", "messageHeadline": "enable BN training in cpu inference build (#8269)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 18, "deletions": 31}}, {"node": {"oid": "996a98b3ac96698f632d0f6fea5d7e6d7d094693", "messageHeadline": "fix the shared provider test for training build; expose more symbols \u2026", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 16, "deletions": 28}}, {"node": {"oid": "b46310b34996bcc47ca73d8ffe7c5af7ebc9a67f", "messageHeadline": "Integrate onnxruntime-extensions into onnxruntime. (#8143)", "author": {"name": "Zuwei Zhao", "email": "4123666+Zuwei-Zhao@users.noreply.github.com", "user": {"login": "Zuwei-Zhao"}}, "additions": 164, "deletions": 0}}, {"node": {"oid": "f616cd07b41d3e216e4400c6b9d151fb2b7d0c14", "messageHeadline": "Provide torch module interface for ORTModule (#8148)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 442, "deletions": 140}}, {"node": {"oid": "ce9d13495284ae77435738fe65f3db349268bec2", "messageHeadline": "gather elements optimization (#8154)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 35, "deletions": 36}}, {"node": {"oid": "ef8f50c4abcd3c53061685f3d2c2def7c1072293", "messageHeadline": "ScatterNDGrad (#8261)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 121, "deletions": 2}}, {"node": {"oid": "97f1eea2ea8e068dd29bd3fefa6043bc067f7eb4", "messageHeadline": "Propagate ROCM version to onnxruntime wheel package (#8247)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "665ecdf9ced7e6ae81ff1fdbf4005d0acc4e0e85", "messageHeadline": "[CoreML EP] Use partitioning utils in CoreMLExecutionProvider::GetCap\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 48, "deletions": 145}}, {"node": {"oid": "4993680e567c2ab9147b155854a547aca5c16972", "messageHeadline": "Graph::GetNodeProvidesGraphOutput -> NodeProducesGraphOutput (#8243)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 37, "deletions": 36}}, {"node": {"oid": "b3479367cfaeabab47737be8cd7df564b3caa379", "messageHeadline": "Add helper to check if node provides a graph output.  (#8186)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 71, "deletions": 60}}, {"node": {"oid": "17d4545ccb85d3480ff35f5676fe00f87e9ca4bf", "messageHeadline": "Improve readability of Graph::PerformTopologicalSortAndCheckIsAcyclic\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 29, "deletions": 27}}, {"node": {"oid": "9b19241b27507aabc10a1d08e15e9865aac3939f", "messageHeadline": "Disable update database for Android code coverage (#8182)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "fa8768723ac690fffa4d2d457c6d96fe55906ae1", "messageHeadline": "Allow custom loaders for testing (#8150)", "author": {"name": "Ankur Verma", "email": "31362771+ankurverma85@users.noreply.github.com", "user": {"login": "ankurvdev"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "507d97b200b09cbe081717f4b3fc8a1996fbeecb", "messageHeadline": "Add initializer for embed layer norm unit tests. (#8196)", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 62, "deletions": 35}}, {"node": {"oid": "9ec0fd6a1cd25ef956542abb7554b0f890e76db5", "messageHeadline": "Revert the cuda algo finding change as this causes a significant memo\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 14, "deletions": 62}}, {"node": {"oid": "83be3759bc939fc0bb5062bc6625d5a1f1312aa0", "messageHeadline": "Add post-install command to build PyTorch CPP extensions from within \u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 333, "deletions": 174}}, {"node": {"oid": "25db5706bb11c00b7244c5a951729cf4fcd160d9", "messageHeadline": "Change \"Export PyTorch CustomOp\" build pipeline to use Ubuntu 20.04 (\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "32ceaf4532a0a47515452cb54683dda786b322cc", "messageHeadline": "POWER10: Optimized SGEMM in MLAS (#8121)", "author": {"name": "RajalakshmiSR", "email": "58744256+RajalakshmiSR@users.noreply.github.com", "user": {"login": "RajalakshmiSR"}}, "additions": 874, "deletions": 405}}, {"node": {"oid": "9b75be3d3e7fc19ae1abfc9e64737b2a8863ab21", "messageHeadline": "Fix a warning in pool.cc (#8168)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "821492f6f5cc75b89479fa5303926bb7c171dc74", "messageHeadline": "Drop std::count_if() in *EmbedLayerNorm Ops. (#8161)", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 16, "deletions": 10}}, {"node": {"oid": "523db6ef443b0a41de1401004ae4b394dad3314c", "messageHeadline": "Check for null runoptions in Run (#8163)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "588511d6da49b0b17727258b5dbc50e7546731d8", "messageHeadline": "Rename embedlayernorm_op_test.cc to embed_layer_norm_op_test.cc (#8160)", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 0, "deletions": 1}}, {"node": {"oid": "800b62a139733f6ec51903136b8ba86a6d7ca80e", "messageHeadline": "Create a quantized EmbedLayerNorm for ORT. (#8124)", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 1262, "deletions": 524}}, {"node": {"oid": "9366114028a80b7467f4714a829a40e05cc05451", "messageHeadline": "make pipelines to support torch1.8.1 and torch1.9.0 (#8084)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 372, "deletions": 420}}, {"node": {"oid": "c716b56f26215788e35aca04644b51a511a4c2bb", "messageHeadline": "Update C++ Standard from 14 to 17 (#8041)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 513, "deletions": 591}}, {"node": {"oid": "9618b6ba6238824cce4176c64efb25bcabfd7117", "messageHeadline": "Fix mac shared_provider warning (#8153)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "a41d0db43cb49d598eb3455ad1dc111f21211d9b", "messageHeadline": "Enable C# GPU tests in Windows GPU CI pipeline (#8142)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "91075255a768814e0524db196ecc33b5c4e8fb2b", "messageHeadline": "Enable TRT provider option configuration for C# (updated version) (#7\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 811, "deletions": 5}}, {"node": {"oid": "49938cce77ca20ddfc1a395c0574e581cb6b121c", "messageHeadline": "Fix Python Cuda loading issues (#7939)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 61, "deletions": 90}}, {"node": {"oid": "378a98597e2c61ccd11343906bd8d83d1af6b7c7", "messageHeadline": "Use std::make_reverse_iterator directly", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 10}}, {"node": {"oid": "00e44861c5ad2f5b746f7313f1d742ec7369527f", "messageHeadline": "Fetching frontier tensors to frontend for ORTModule (#8086)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 472, "deletions": 1}}, {"node": {"oid": "eb36258df4764a73d6af5b82c05f24124775cd99", "messageHeadline": "Enable signed int8 data type for activations in static quantization (\u2026", "author": {"name": "SilvanK4t1qbit", "email": "48225758+SilvanK4t1qbit@users.noreply.github.com", "user": {"login": "SilvanK4t1qbit"}}, "additions": 200, "deletions": 48}}, {"node": {"oid": "e083d207cfaf0612b08be503383e29efe7646f82", "messageHeadline": "Disable InitProvidersSharedLibrary when training is enabled. (#8132)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7ed9f5fc90665fd94aeb8f2900806716b97b9c11", "messageHeadline": "[Java] Fixing the creation of OnnxTensors from scalars, adding tests \u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 167, "deletions": 28}}, {"node": {"oid": "80b7b134bf33ccde32d7a40366c7f1848b0a3516", "messageHeadline": "Adding optional ops in contrib ops (#7946)", "author": {"name": "Negin Raoof", "email": "neginmr@utexas.edu", "user": {"login": "neginraoof"}}, "additions": 464, "deletions": 8}}, {"node": {"oid": "59e336040c505afe5ef042357c18262734d0f975", "messageHeadline": "Ortmodule override torch.manual_seed() (#8131)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 20, "deletions": 0}}, {"node": {"oid": "b478086bc1ca26070e61d7277cdd91d9d9b0eb76", "messageHeadline": "Fuse attention node even in case of different Q,K hidden dimensions (\u2026", "author": {"name": "Viswanath Boga", "email": "44417868+viboga@users.noreply.github.com", "user": {"login": "viboga"}}, "additions": 173, "deletions": 62}}, {"node": {"oid": "4fd7efcf0d891ccc9a66ced0b1ae8c0c22f1928a", "messageHeadline": "Update logic in props.xml to account for shared provider library chan\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 40, "deletions": 3}}, {"node": {"oid": "f000dfddbe43ef753dc8355fc16e98c9bb680f87", "messageHeadline": "Update run_dockerbuild.sh: set default python version based on OS ver\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "1fa69866563f99edce9dc4be8b17ab835928de9b", "messageHeadline": "Chang how numpy version is handled. (#8130)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "db88f3059c6a16e35326992fc0f1f73b0a020771", "messageHeadline": "[js] fixing broadcast issues in pack mode (#8090)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 189, "deletions": 37}}, {"node": {"oid": "cbdd59dae9c091a2482d2d8cda42163f7447c3ec", "messageHeadline": "MLAS: enable SSE 4.1 path for x86 build (#8127)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 6, "deletions": 8}}, {"node": {"oid": "45ce239929dbb8cc9fcfe3e5a2dc3fe5ad5c80ae", "messageHeadline": "User dynamic axes in one step beam search output (#8092)", "author": {"name": "Xiaoyu Liu", "email": "lixiaoyu@microsoft.com", "user": {"login": "xi-liu-ds"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "cccd61e3bc6f15120932a032a48bbdcc569938ea", "messageHeadline": "Add int64 as a required type to ConstantOfShape as it's used by the p\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "b1e21312b5565c00bf39ad5ff8107657100b5204", "messageHeadline": "[Mobile package] Update required operator config with additional ops \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "664e548e318944be2d1bd46ecc4caf7139b237b6", "messageHeadline": "Col2im optimization by eliminating integer multiplications:", "author": {"name": "Evgenii Indenbom", "email": "eindenbom@microsoft.com", "user": null}, "additions": 180, "deletions": 74}}, {"node": {"oid": "6e2b064aec3ab2dd5d6d5a92d234f8c7d1afe52e", "messageHeadline": "Delete some unused code in run_dockerbuild.sh and Enable Nuget CUDA t\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 106, "deletions": 310}}, {"node": {"oid": "f6292d9b387b34a53d2b47d98a44d7195a324d0f", "messageHeadline": "[Android] Output error message to android log instead of stderr (#8114)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 73, "deletions": 43}}, {"node": {"oid": "9003df5d872c0196494308cd210e5e814d664d53", "messageHeadline": "Fix 32bit Android java API crash (#8122)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 10, "deletions": 5}}, {"node": {"oid": "4bb0e29d0ea9565f2118fa0686806de6a595c98b", "messageHeadline": "initialize generated_value_names with graph input (#8085)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 13, "deletions": 2}}, {"node": {"oid": "839f69d249a616972250c86201df296231ff7770", "messageHeadline": "Implement WINRT_IMPL_LoadLibraryW to avoid calling LoadLibraryW direc\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 5, "deletions": 0}}, {"node": {"oid": "e7d7fa8fa2c85f5b34bfb4c67f0dc181c5d43dc1", "messageHeadline": "Update migraphx to rocm4.2 (#7994)", "author": {"name": "Shucai Xiao", "email": "shucai@gmail.com", "user": {"login": "scxiao"}}, "additions": 100, "deletions": 47}}, {"node": {"oid": "5809890ba23af1cb4f355ff7793c9e0f7d1c58fb", "messageHeadline": "Fix a compile error in InferenceTest.cs (#8119)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "8cacb26946030449bb6ea6b079a2d0a6f0ff8697", "messageHeadline": "remove debug.keystore from repository due to a credential issue repor\u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "27d1784d44826e1f1b58b76dadce88617e5f1f4d", "messageHeadline": "Add TRT 7.1 Pipeline (#8073)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 298, "deletions": 5}}, {"node": {"oid": "3cd06cb38cd5e76619dbd7a2859897ab9cdd2d8c", "messageHeadline": "Added support for ReduceMean on DNNL EP for CPU and GPU (#7902)", "author": {"name": "chethanpk", "email": "63478277+chethanpk@users.noreply.github.com", "user": {"login": "chethanpk"}}, "additions": 271, "deletions": 15}}, {"node": {"oid": "352d560fd586c35f96495092e75234b644ec1d8e", "messageHeadline": "Adding Conv+Clip fusion (#8102)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 71, "deletions": 15}}, {"node": {"oid": "10b7ed6430feba144b77d4365bc2129af6d5d5c5", "messageHeadline": "Added op_name to message when we are missing a kernel. (#8110)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "cba4bc11c78a55fa3aeb7c1490e8f9b387dceeec", "messageHeadline": "Split Linux CPU CI pipeline (#8097)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 61, "deletions": 7}}, {"node": {"oid": "51c12a715b8b7ab91dc13752e257fd4ec4eba236", "messageHeadline": "Add NGramRepeatBlock contrib op (#8078)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 427, "deletions": 0}}, {"node": {"oid": "5ac06bad619345a1a7d760c7d69666b31e796372", "messageHeadline": "Relax test tolerance to make CI more reliable (#8100)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "059d705988af5f836968d5566ff51bcc052c64fa", "messageHeadline": "support pass in custom op registry for eager mode (#8087)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 81, "deletions": 5}}, {"node": {"oid": "9f5969693ac654c1b33d84e2684681d412b0f354", "messageHeadline": "clean up builds for interop_torch (#8017)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 578, "deletions": 468}}, {"node": {"oid": "5c2e1bbb0a3bd2f3908e56b99779dc28035e6d4e", "messageHeadline": "Fix input schema extrator for ORTModule (#8098)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "7701c8703e98215e2c49e6fd3025f91e90cabd1e", "messageHeadline": "Add module attribute to ORTModule to support HuggingFace Trainer save\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 49, "deletions": 3}}, {"node": {"oid": "08eeb8763dd3aeaaa8b1bc31e9327d15d4849b33", "messageHeadline": "Loosen validation checks in Concat to unblock execution of model in #\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 107, "deletions": 46}}, {"node": {"oid": "b2247ece25b335429c3c6c73f4935a9b6d2c3574", "messageHeadline": "Make Perf Test Configurable (#7836)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 364, "deletions": 456}}, {"node": {"oid": "aa68157c3ddc28779d5bb56c0ca8e51f94f598dd", "messageHeadline": "[Mobile package] Update required operator config with additional ops \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "d83f7fd4aa176f56d9093e19a1db13399aac0c8a", "messageHeadline": "[NNAPI EP] Enable Slice support (#8031)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 542, "deletions": 224}}, {"node": {"oid": "96989b83ee2da94e9eae78921d7ef0ee507e40ce", "messageHeadline": "Create python packages for DML (#8061)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 118, "deletions": 34}}, {"node": {"oid": "d924fd205b98e85e1ffd9025b842d65506a81dd8", "messageHeadline": "Create and move quantization tests to a shared Quantized utils file. \u2026", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 109, "deletions": 76}}, {"node": {"oid": "365070b74400b022182b505bf24b1c6db6cacdcb", "messageHeadline": "Merged PR 6158327: RI Onnxruntime github into DmlDev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 6384, "deletions": 1651}}, {"node": {"oid": "32ef39be58716c0ac9529e01957f96808c138312", "messageHeadline": "[Android] Move add header files into AAR to using Gradle (#8068)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 76, "deletions": 73}}, {"node": {"oid": "1d8edd0b5bcf3c3172cc10ae60662073e8f8f748", "messageHeadline": "Fix missing files on linux (#8066)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "c76172fab664f20f7b173b80e1a07f66bb955f7d", "messageHeadline": "Fix PythonOp with input which has no gradient (#8011)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 72, "deletions": 23}}, {"node": {"oid": "de8f2ecda90ca7d7199ae7a6dd00c2408bc0bdc8", "messageHeadline": "Reduce Kernel Optimization (#8067)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 75, "deletions": 28}}, {"node": {"oid": "0ebaa71f49794a849cd9a625ef8c55747a64eda2", "messageHeadline": "Improve Windows Platform system error messages (#8063)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 19, "deletions": 20}}, {"node": {"oid": "32e118bef023d4083a6f3e3c99e867605d1237b7", "messageHeadline": "Fix microbenchmark build failure (#8064)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e31784b6cffc591dbe99165e0a21a6aaefd34426", "messageHeadline": "decouple the python module construction from pybind_state (#8060)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 17, "deletions": 1}}, {"node": {"oid": "96cf533c76ef48db842f91b86740482fe85a7146", "messageHeadline": "Remove DML from Windows GPU CUDA 10.2 pipeline", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "25c49a5fe0e8ab7f9dca88678973b354f06901cb", "messageHeadline": "fix issue with cmake path (#8055)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "07b166bb1bc40714230d3405a08d52b6aee53fb3", "messageHeadline": "fix PATH addition in windows", "author": {"name": "iperov", "email": "lepersorium@gmail.com", "user": {"login": "iperov"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "887c3149e35ecd92a764c45b890824dfa651f008", "messageHeadline": "[js/react_native] Use a mobile ORT instead of a full ORT (#8042)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 324, "deletions": 553}}, {"node": {"oid": "6a1b0001251f4918c31ef68df76c49c9097d1dd9", "messageHeadline": "Fix unit test typo in test_op_embed_layernorm.py (#8056)", "author": {"name": "Nick Kreeger", "email": "nickkreeger@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "07788e082ef2c78c3f4e72f49e7e7c3db6f09cb0", "messageHeadline": "Enable python GPU tests (#7854)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 23, "deletions": 34}}, {"node": {"oid": "de1b1edcad962552da09f14433e484a42c7a6525", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 6384, "deletions": 1651}}, {"node": {"oid": "8079c763835a594b3c76249398f92c3372901f23", "messageHeadline": "Create ORT opschema library (#7903)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 166, "deletions": 12}}, {"node": {"oid": "c72a8c7ff417a3277b4d1812e8cd7e8e853050ec", "messageHeadline": "Upgrade tf 2.4.1 to 2.4.2 for component governance (#8036)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9acf93b90af1caca95fc6b28ebd2374d06bd94c9", "messageHeadline": "Take graph topology into account when creating dnnl subgraphs (#7910)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 29, "deletions": 4}}, {"node": {"oid": "6d7461795f4e9e0a4a5280f14a87a1828c66b6ff", "messageHeadline": "Update Version.md (#8021)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ad6a306a7fae662ef157dcc3954aeb61d1c6be2b", "messageHeadline": "Add pragma once (#8040)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "96ead2be916174118d20612e2ac54da667dcaa77", "messageHeadline": "Avoid hashing the operator type in the GraphViewer priority node chec\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "6e134c2cc31f62811cc142e45d8320108031b7a0", "messageHeadline": "[Objective-C API] Add support for documentation generation (#7999)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 63, "deletions": 23}}, {"node": {"oid": "1d7f44a8329fbf8dc0fc2e25e603d439998ce138", "messageHeadline": "Add unit test for EmbedLayerNormalization quantization op. (#8033)", "author": {"name": "Nick Kreeger", "email": "nick.kreeger@gmail.com", "user": {"login": "nkreeger"}}, "additions": 144, "deletions": 0}}, {"node": {"oid": "e6225c62a5337e6e42d90fc4303f570dcef3bb3e", "messageHeadline": "transformers test CI pipeline fix (#8016)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 17, "deletions": 4}}, {"node": {"oid": "43c45ddd66f4ee01d7d6f63b785423cdf874aadd", "messageHeadline": "Update DirectML EP changes from DmlDev as of 2021-06-07 (#7987)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 7, "deletions": 32}}, {"node": {"oid": "8b0c2e1f3d72db6708317d6c655e7ee68e961cba", "messageHeadline": "Merged PR 6101363: Int64 prototype work for ONNX runtime DML EP", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 122, "deletions": 76}}, {"node": {"oid": "2f2aaf2cf6151e75c4023fc2d0c9d044f5cd212c", "messageHeadline": "Fix Memory Leak from DlpackToOrtValue (#8029)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "d02de9c1bcf0eda46c50cb280440007b04218baf", "messageHeadline": "[ROCm] dockerfile updates (#7955)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 36, "deletions": 1}}, {"node": {"oid": "00d48d9c30051f279a99f70e38191adb782f2bbf", "messageHeadline": "Add enhanced partitioning utils for use by compiling EPs (#7991)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1027, "deletions": 112}}, {"node": {"oid": "35ca3c99d159bdf8d042a4f2a28dfb5ad5e21d42", "messageHeadline": "Fix ROCm wheels pipeline after changes to manylinux scripts (#8026)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 73, "deletions": 22}}, {"node": {"oid": "20579595c8861a79c09953784acfcaa69c436559", "messageHeadline": "Make logic in InsertCastTransformer around forcing a node to fp32 mor\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 263, "deletions": 65}}, {"node": {"oid": "02372251174fa0721fad106e411860b4a337196e", "messageHeadline": "Add @file annotation to support doxygen generation of C API docs (#7458)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "b2ed4fb0a48ef7e0444011b4f97b53cac7ba1f65", "messageHeadline": "Merge orttraining and ortmodule single gpu ci pipelines (#8022)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 56, "deletions": 67}}, {"node": {"oid": "4d1b48632cd86a9155834056e6ee4c276e590dc9", "messageHeadline": "[CoreML EP] Add ArgMax op support and modify OpBuilder interface (#7924)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 351, "deletions": 32}}, {"node": {"oid": "b313c4581cba770c6e7e853390525c0f38e98d89", "messageHeadline": "Remove CC/CXX env settings from C API packaging pipeline (#8014)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "2a74f5e85baedb9b5e52a601d14e0e8974a4776d", "messageHeadline": "Save module output for backward if needed (#8010)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 97, "deletions": 0}}, {"node": {"oid": "c74265667e1cbe48d39598f71b9b0534869d9b38", "messageHeadline": "Remove CUDA architectures 35 and 86 from GPU packages (#8004)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "b03383f6d59a1c7352f4ccb252c61cfb41220f9d", "messageHeadline": "Add cuda provides files (#8002)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "f013b0c0eb819685c3c2bdc8a9886d3597be15c6", "messageHeadline": "[NNAPI EP] Add support of Elu, merge in NNAPI updates for API level 3\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1182, "deletions": 149}}, {"node": {"oid": "aa45545af7ef580735d06dcc7c0dbecae5b4682f", "messageHeadline": "Update orttraining-linux-gpu-perf-test-ci-pipeline.yml (#8005)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "cb5f411da3ca5d05002c1d44f9519df3dbffd37c", "messageHeadline": "Fix Python Packaging Pipeline && Build Clean Up (#7993)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 76, "deletions": 109}}, {"node": {"oid": "d433aa2459a80b4d53715f29779cc3b8ead2e37b", "messageHeadline": "Add transformers tool test to pipeline (#7959)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 123, "deletions": 574}}, {"node": {"oid": "f0f301266646f5942cdac00270af9b74ffe371f7", "messageHeadline": "Add SoftmaxCrossEntropyLossInternal to Support Dynamic ignore_index I\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 863, "deletions": 53}}, {"node": {"oid": "f38200e209206212866792337c95b17fc0cc8553", "messageHeadline": "Override ORTModule named_modules to support extra arg (#7954)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "1cc896c8ae9cccac81d05e6b4adf52d55427c9f2", "messageHeadline": "optimize js package folder structure (#7989)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 44, "deletions": 38}}, {"node": {"oid": "47d8977741715391a7a829fc12387211928db022", "messageHeadline": "add missing provider_options.h in packages (#7995)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 10, "deletions": 60}}, {"node": {"oid": "275796a16579d77f13699ce8c3e21179e0e8629b", "messageHeadline": "Update googletest to latest commit to fix build issues with GCC11 (#7\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 8}}, {"node": {"oid": "861cd0fb249980497e03364823631dbc4139e55f", "messageHeadline": "Increase Python Mac Job Timeout (#7998)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "500f18badbb36e30d40f6dab9c4bf0fe277c60b8", "messageHeadline": "fix bug that bias can not be shared across Convs (#7982)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 114, "deletions": 89}}, {"node": {"oid": "f5c6b9703a65548fe603405b32a3f7e0b8e0d108", "messageHeadline": "Merged PR 6129591: Merge github master into Dmldev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 14768, "deletions": 2223}}, {"node": {"oid": "66170bfceff693337d172761ca2fe76b1e8766e9", "messageHeadline": "Python with DmlExecutionProvider : choose device_id in SessionOptions\u2026", "author": {"name": "iperov", "email": "lepersorium@gmail.com", "user": {"login": "iperov"}}, "additions": 12, "deletions": 1}}, {"node": {"oid": "4ecbae43b2b4f94f2d3c82e6cf8a04042c441529", "messageHeadline": "Use GCC 10 in Linux CPU CI pipeline (#7985)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "a776b57160e3ca937573248349a9377c90e2bf24", "messageHeadline": "Add shape inference to custom symbolic functions (#7937)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 12, "deletions": 7}}, {"node": {"oid": "dce76c15e7c6c20b5455e05b58505c8e4d3e7129", "messageHeadline": "add dockfile for ROCm 4.2 (#7749)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 171, "deletions": 0}}, {"node": {"oid": "b50e9d9d74b2ff70277ba0e8fcaa8ba157944664", "messageHeadline": "Adding webgl shape kernel (#7971)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 42, "deletions": 1}}, {"node": {"oid": "0f01de3b0baf76447f2d15f8c004f0007f338b17", "messageHeadline": "[js/web] Add wasm SIMD backend to onnxruntime-web (#7896)", "author": {"name": "Gao, Chun", "email": "53756924+chun137@users.noreply.github.com", "user": {"login": "chun137"}}, "additions": 139, "deletions": 65}}, {"node": {"oid": "71c4f5ddb2181b76f58442193c210f891aa5c085", "messageHeadline": "ATenOp Enhancement (#7725)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1004, "deletions": 123}}, {"node": {"oid": "0696e2f0d49e015262d4d03e2f648737da2af418", "messageHeadline": "[Objective-C API] Add script to assemble pod package files. (#7958)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 212, "deletions": 10}}, {"node": {"oid": "fc331cbd5d2c8260298d6f0071987dc5af665f6a", "messageHeadline": "Accept success codes other than S_OK in RoInitialize (#7979)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "8c951b94d1daa7eda49cf6645730d6b58e9da2ed", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/rylai/ri_6_4", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 11018, "deletions": 652}}, {"node": {"oid": "eb354853d34cef0b683972fe7c4188e6309fad02", "messageHeadline": "Update CMakeLists.txt for openvino EP (#7980)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "1a5ee11dbd0dbf81977e222d243a720cbc8c5672", "messageHeadline": "Implement Sequence Ops GPU (#7863)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 477, "deletions": 42}}, {"node": {"oid": "9e4dc08483fe73bfc299054d841af48704e190e8", "messageHeadline": "training with custom autograd Functions (#7513)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 3681, "deletions": 141}}, {"node": {"oid": "4cb3c5e3e2c165755ce0c74630241f505c67df1a", "messageHeadline": "Update auto-generated csharp files (#7950)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2902, "deletions": 54}}, {"node": {"oid": "e23529f313d401d62e97f0d63efe61391bf7c810", "messageHeadline": "Update Python Wheel File Path to fix python packaging pipeline (#7978)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "bfa996b5fa1ce87d2501abd8371ec5612d288cac", "messageHeadline": "add emsdk to component detection ignore dir (#7932)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "429df40f1dfa0cd72b2fc449f80c7945c79a4b07", "messageHeadline": "Suppress warnings in GTest (Fixes Gcc11 build errors)  (#7957)", "author": {"name": "Ankur Verma", "email": "31362771+ankurverma85@users.noreply.github.com", "user": {"login": "ankurvdev"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "9b5f7491767d0151da3f01effaa2e9bc08adff47", "messageHeadline": "[wasm] emsdk: allow to install emscripten only (#7961)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 3288, "deletions": 30}}, {"node": {"oid": "f352d54743df1769de54d33264fcb7a2a469974d", "messageHeadline": "Refactor Resize/Upsample implementation to reduce binary size.  (#7650)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 534, "deletions": 355}}, {"node": {"oid": "fd23b8caaddc4c460463774d696af13bef63aa46", "messageHeadline": "Update mobilenetv2 quantization notebook (#7941)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 81, "deletions": 14}}, {"node": {"oid": "291453dac94f0aaf42df5b776158139c863bc064", "messageHeadline": "[wasm] fix test report generation (#7953)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 40, "deletions": 6}}, {"node": {"oid": "09ab8955634d98086915b1124db7a97dce46e05c", "messageHeadline": "Fix a resolving issue on the quantization node transforming. (#7952)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "64f1a4ed224d45432ebb5104a2e4f7f986de761a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 3762, "deletions": 1583}}, {"node": {"oid": "b856e7ae3cb135508425af89f29a71888b2c0150", "messageHeadline": "Update build.py: change default cmake generator for Windows to VS2019\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0975e7c9a71a4c4060ba062492799a4f90dfd311", "messageHeadline": "Add and correct multichannel test for qlinear pool test. (#7864)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 22, "deletions": 2}}, {"node": {"oid": "5a7f65b831286c7a095d9f56ecb54b7f0823f9bd", "messageHeadline": "Fix training e2e pipeline (#7942)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 44, "deletions": 43}}, {"node": {"oid": "2b3f953701310a4f604d6f93250e9ab4434a0802", "messageHeadline": "Run PadGrad UT for CUDA Only (#7947)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "650314c9260e5852333112838ada806159dbe473", "messageHeadline": "[wasm] bugfix for emscripten '--preload-file' under node.js (#7944)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "0723d164366e274add60b09360105d5b8bace633", "messageHeadline": "[wasm] allows to specify MALLOC setting for wasm build (#7934)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "6a9023f47d59637c7eb0b74a538a83d4b2c6e3ed", "messageHeadline": "[JS/Web]Adding support for WebGL v1 (#7906)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 12, "deletions": 83}}, {"node": {"oid": "ab973dce339ad895b0d3c94032d11045118dbe09", "messageHeadline": "[Objective-C API] Enable CoreML EP (#7914)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 144, "deletions": 5}}, {"node": {"oid": "ed5fd919ef68ac29bafbd2fbdc436226f40448d3", "messageHeadline": "Update dockerfiles to use the latest cmake (#7933)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 39, "deletions": 47}}, {"node": {"oid": "8f8b9302a20f50877c223e63a2e63d5750fdd27e", "messageHeadline": "Re-enable some of the recently disabled cuda tests (#7873)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 66, "deletions": 97}}, {"node": {"oid": "a118da160d591eba7040cdd9798a06794e1d41c3", "messageHeadline": "pad gradient (#7926)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 88, "deletions": 1}}, {"node": {"oid": "3bb780dcd5a50b085bc5917f8b4de9980eecfbc9", "messageHeadline": "Update Vitis AI EP to support multiple DPU targets through provider o\u2026", "author": {"name": "Jorn Tuyls", "email": "jtuyls@users.noreply.github.com", "user": {"login": "jtuyls"}}, "additions": 147, "deletions": 47}}, {"node": {"oid": "896f32ec0991224c9f33ef3c5db61bc57f6e76f3", "messageHeadline": "[js/web] support string tensor for wasm backend (#7891)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 195, "deletions": 94}}, {"node": {"oid": "b854f2399dafb379584a0d7aebe42c958ae5aede", "messageHeadline": "Update manylinux build scripts and GPU CUDA version from 11.0 to 11.1\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2639, "deletions": 1089}}, {"node": {"oid": "a272a75cd1a926ef17aa4bd2a550921e1b0dce80", "messageHeadline": "[js/web] allow pull wasm artifacts from CI (#7886)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 197, "deletions": 21}}, {"node": {"oid": "2fd703a97af5d82112cf2b573232cf69bf4ae96c", "messageHeadline": "Merged PR 6115647: RI github into dmldev. Need to merge this unsquashed", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "5c8b07577d3767496acde671ca0de38c6d4cfae7", "messageHeadline": "Merge commit '0fbec1b9c1d8f34a44ac5e7efc1fee0a0a08ac84' into user/ryl\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "14fe3e8404f94a40fb5bba4277ecf63f6e66d447", "messageHeadline": "Merged PR 6115559: RI changes from Github into dmldev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 8414, "deletions": 2461}}, {"node": {"oid": "4f82ad1b58a3d2c671fb64604b03108d9c295b7a", "messageHeadline": "Topo sort the model before saving (#7913)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 63, "deletions": 1}}, {"node": {"oid": "c45ac166d36b28691962e215a3ca30343156d44b", "messageHeadline": "Add graphviz into Dockerfile images for Python API documentation (#7819)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 46, "deletions": 20}}, {"node": {"oid": "faae347d9ff40a03b1b32117ded1ed717d3b536e", "messageHeadline": "[wasm] upgrade emsdk version to 2.0.23 (#7893)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 10, "deletions": 12}}, {"node": {"oid": "79a6727a020e6ece3d8a15f1456f8f167b291fa6", "messageHeadline": "Add podspec template for ios package, update build settings (#7907)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 35, "deletions": 13}}, {"node": {"oid": "0e935b8718a3d2e6316baff4ba9067bd382fa746", "messageHeadline": "Fix typo (#7872)", "author": {"name": "Billy Franscois", "email": "69786531+BillyFrcs@users.noreply.github.com", "user": {"login": "BillyFrcs"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "271a34302406fd14a138fa39b03d21335948d0ad", "messageHeadline": "Cache initializers and avoid device check ot end of forward (#7905)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 8, "deletions": 7}}, {"node": {"oid": "9946e6f7dfe898bb73429d633da4639cd92bc5ae", "messageHeadline": "fix broken tests (#7909)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "0fbec1b9c1d8f34a44ac5e7efc1fee0a0a08ac84", "messageHeadline": "Update the operator documentation generation (#7787)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 563, "deletions": 854}}, {"node": {"oid": "3d734a1cdcd3c45cf8f37302e70d58ca6e56f18a", "messageHeadline": "Missing logic for cuda nuget package (#7911)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "38ca0f48398f10b6eae745988c7a1ebd276bbba5", "messageHeadline": "Change CMAKE_CUDA_STANDARD to C++17 for Windows GPU build (#7883)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "a9f7eef75476c7e3cf18ee83a75183e5f274c9f2", "messageHeadline": "Add API_IMPL_* blocks around shared provider methods as they are C AP\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 16, "deletions": 1}}, {"node": {"oid": "f9587d6051213c143d271396230dd58601aed778", "messageHeadline": "[js/web] update README.md (#7894)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "6d9062641c020c52f888a07b8d85fc6678cab66d", "messageHeadline": "Basic data parallel tests for ORTModule (#7812)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 169, "deletions": 0}}, {"node": {"oid": "e7e200ee59df07e3eb3aeb8220171ef72fcd1a4e", "messageHeadline": "Add test for iOS package (#7816)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1084, "deletions": 10}}, {"node": {"oid": "81ed6c55bf3fb7e0e5f85ccf2e71092eeefcf3d7", "messageHeadline": "fix grouped pointwise convolution (#7885)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 42, "deletions": 20}}, {"node": {"oid": "3a72932c4a4afc2e72c6b3bfc17b3c0a248e6232", "messageHeadline": "Don't hold onto unnecessary numpy references while binding numpy obje\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d8bcb3d6a4eaef5b01300a96a8746a8b6f0907f4", "messageHeadline": "Added virtual destructor to adasum_interface.h (#7882)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "451fcb7df1185d3a6d51fed9fa28c14622b9420c", "messageHeadline": "Add sequence support for identity on GPU (#7810)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 173, "deletions": 46}}, {"node": {"oid": "bccf29210c25558d7999de93eb3540869bf1b994", "messageHeadline": "Merged PR 6103324: Remove usage of non-generic error code (FWP_E_NULL\u2026", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal@microsoft.com", "user": {"login": "sumitsays"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4dd724ef1ad5d15a9bc56541e494dbbc03deca4b", "messageHeadline": "Enable WebAssembly SIMD build (#7839)", "author": {"name": "Gao, Chun", "email": "53756924+chun137@users.noreply.github.com", "user": {"login": "chun137"}}, "additions": 31, "deletions": 11}}, {"node": {"oid": "5a63904aa9d089a47b313d52d5f95c7318c54d87", "messageHeadline": "Remove some templated versions of functions that are no longer needed\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 20, "deletions": 41}}, {"node": {"oid": "3f43a84e10776fada10c94ea68ad2334b09ddc58", "messageHeadline": "Merged PR 6093117: Fix test_DynamicQuantizedLinear_max_adjusted_expan\u2026", "author": {"name": "Sumit Agarwal", "email": "sumitagarwal@microsoft.com", "user": {"login": "sumitsays"}}, "additions": 7, "deletions": 32}}, {"node": {"oid": "e41e042de6b6ef7bb9cf88c75533c40a769179eb", "messageHeadline": "[OpenVINO-EP] Adding OpenVINO-EP samples to Msft Repo (#7826)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 1874, "deletions": 0}}, {"node": {"oid": "ab4b5055c7411c4dc835bfd1ffb91190956291d8", "messageHeadline": "[Objective-C API] Fixes from package testing and clean up (#7866)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 116, "deletions": 86}}, {"node": {"oid": "35b49b64c72405c22720d325a0f5f36379b271f5", "messageHeadline": "Fix regex to detect Objective-C/C++ (.m/.mm) files. (#7870)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0255c83dc4b93e3d15b83363a523348b8db38b16", "messageHeadline": "Clean up CPU kernel definition for opset 13 Pad (#7867)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 5}}, {"node": {"oid": "71b05f74a2d26379129f69b842b1ea0ae09127f6", "messageHeadline": "fix duplicated node name (#7865)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 32, "deletions": 19}}, {"node": {"oid": "1f4421fe70fb96f67d2fd5b098b90c171d5ef273", "messageHeadline": "Include ORT C/C++ API headers in the ORT Mobile AAR package (#7858)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 65, "deletions": 15}}, {"node": {"oid": "63df683040e28329f74b1adad78235fba24f47be", "messageHeadline": "Fix path used in check for cudnn library (#7786)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "ddf4aaaae1c6ad214f146a2cbd8d5b4bec818013", "messageHeadline": "Resolve issue with wrapped ORTModule load_state_dict (#7847)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 127, "deletions": 36}}, {"node": {"oid": "8140e3fde5c22d8fc95a50fa2f840fa15238daba", "messageHeadline": "Make requantize a qgemm post processor (#7850)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 228, "deletions": 195}}, {"node": {"oid": "ccdedf1b2e0dca5e64eb0016efccaef2d6922bae", "messageHeadline": "[js] update documents (#7852)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 451, "deletions": 58}}, {"node": {"oid": "738021971738d99ca28cef2fe4e8e0af29cd09e4", "messageHeadline": "Fix bug in Transpose CUDA kernel (#7329)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 190, "deletions": 62}}, {"node": {"oid": "883923a40acbe40e29c5a0da7365e9c8cab25e8c", "messageHeadline": "Support bool type for Pad CPU (#7856)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 80, "deletions": 21}}, {"node": {"oid": "13622bae91a7f67757a86f16eb48aa810aaf01ee", "messageHeadline": "Add Apple log sink. (#7820)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 146, "deletions": 60}}, {"node": {"oid": "45a7352622b4590fa2787a78be1bb6d5f0224236", "messageHeadline": "Update Mac CI builds to use macOS-10.15 image, Xcode 12.4. (#7437)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "2a3851cd75cd240c6e8e0f50736f3b24cf7a2dc9", "messageHeadline": "fixed bugs in packed mode and enable pack mode tests in ci (#7848)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 74, "deletions": 64}}, {"node": {"oid": "bed6e87cbdbe1120f7958b6e1eeff5eda3c11e8e", "messageHeadline": "add environment variable to control default training package's local \u2026", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 19, "deletions": 8}}, {"node": {"oid": "fa093d8e45c5686f438c93146ab31840dd31f779", "messageHeadline": "[Objective-C API] Add ORTSession methods to get input, overridable in\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 117, "deletions": 1}}, {"node": {"oid": "94bb09bf47d9fa76247e81b8c7765ab344da5f0f", "messageHeadline": "fix topo sort in quant tool (#7833)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 55, "deletions": 14}}, {"node": {"oid": "fc472a04be4a3308affa2184913d354dd8b6043f", "messageHeadline": "Relax tol for Conv1D fp16 test (#7844)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c08bb4eee309fd7382b77a78cec8a2dbd37bda15", "messageHeadline": "Update docs/ONNX_Runtime_Server_Usage.md (#7818)", "author": {"name": "Siva Popuri", "email": "siva.popuri@alumni.stonybrook.edu", "user": {"login": "asamadiya"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "afca89dce6d852dd8faabb5c53e6a9ff739de141", "messageHeadline": "fix boost download url (#7843)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "331f20428c6c94f6c485ab5210a9bf8dc9391f04", "messageHeadline": "[js/web] only apply max thread number when it's omitted (#7834)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "c5ea5907c09cc39077749971637852ab424a5c50", "messageHeadline": "Fix permission error for ORTModule lock file (#7814)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 35, "deletions": 13}}, {"node": {"oid": "a85a9ddaf5f83045ad770964be336ba5894dc5b2", "messageHeadline": "Merged PR 6091402: Fix inbox ORT debug info", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "c487824a31fe11cd96c7fc1c6c0dbefa609c4009", "messageHeadline": "Fix bug in Einsum implementation (#7822)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 389, "deletions": 6}}, {"node": {"oid": "d1f0251e3913934e73cee64398e915032618cb01", "messageHeadline": "Python bindings fix ups in preparation to Sparse Tensor introduction \u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 932, "deletions": 687}}, {"node": {"oid": "57782b3463735402c582721bd1883363ada2fb4a", "messageHeadline": "Add supported operators/types documentation for the ORT Mobile packag\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 211, "deletions": 0}}, {"node": {"oid": "4fe59c8b29613fcd9b8fb6125a05358164f81bb5", "messageHeadline": "delete model_copy to save memory allocated in forward call (#7832)", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "1c6b6f696eae4a1841ece99cea98380de6b1c3f7", "messageHeadline": "fixes for cuda centos/manylinux (#7830)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 25, "deletions": 7}}, {"node": {"oid": "f78af4fc8cf3bc5b6d8a486af78473840f12612f", "messageHeadline": "Use RTLD_GLOBAL for onnxrutime_providers_shared on unix (#7831)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 51, "deletions": 45}}, {"node": {"oid": "ea1a4f8fb5e00839e225853e68661b43d95dbd5f", "messageHeadline": "[JS]support running super resolution model using ortweb (#7677)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 903, "deletions": 41}}, {"node": {"oid": "6ca1ee77332ab89ec60064ea4542fa68fed71b3c", "messageHeadline": "Fix rpath issue with pybind. (#7829)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "29c68888af36941fd0653232a0848704ca7d0f5a", "messageHeadline": "Update BERT convergence baseline.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "3d12e957a7b99bbe41c2344f0dc6d859530e1ba1", "messageHeadline": "Workaround for miopenReduceTensor() behavior difference in ROCm 4.2", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 110, "deletions": 4}}, {"node": {"oid": "f49a4b63298df2c5f890dfea210cfb3c702c2516", "messageHeadline": "Decrease lock contention in qlstm by memory allocation. (#7815)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 48, "deletions": 15}}, {"node": {"oid": "c0a8905b90546e30759ba93e9e2fdb134790ebe5", "messageHeadline": "Merged PR 6087868: Update ORT to latest master", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 6588, "deletions": 4299}}, {"node": {"oid": "3fb4a2dc991b576494dfcf5f364527336f748541", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/justoeck/ri_\u2026", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 6588, "deletions": 4299}}, {"node": {"oid": "ff655175ff2ecb32f24cdc43870a6c34667d73d6", "messageHeadline": "Eliminate no op node - add 0 (#7798)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 221, "deletions": 0}}, {"node": {"oid": "9241d7639656cdf3e3eac84b1aab2165f8b3e1ba", "messageHeadline": "Remove unnecessary cuda libraries refernced in cmake (#7824)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "93c8e29782138b71a9e0605745192e92c0613f2e", "messageHeadline": "Improve code coverage report (#7770)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 33, "deletions": 26}}, {"node": {"oid": "cdd2129fed4c7583011338485143e641084366a0", "messageHeadline": "Fix CUDA Pad kernel registrations (#7813)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 37, "deletions": 14}}, {"node": {"oid": "a41255c280b06a7de748f211af42abccc724a225", "messageHeadline": "Fix performance regression in Reduce operators introduced by PR #7206\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 161, "deletions": 107}}, {"node": {"oid": "f487f6be2577d8b69a1378ff79f71b87c270f3eb", "messageHeadline": "Add int8/int32 Relu for Opset 14 (#7536)", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "98007f0be6b9961bc8703b0ee5109fe71e0d17be", "messageHeadline": "Fix typo in the ios packaging script (#7802)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "02c78a8aa8ec79653ca6c78ed351a6bc0e29cc64", "messageHeadline": "test migration to rocm4.2 (#7800)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "da5ab325eeecf70b08614e27ceb5a3e99cac291c", "messageHeadline": "register batchnorm for opset14 cuda provider (#7806)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 30, "deletions": 7}}, {"node": {"oid": "bd5067a2ff0769b5f11ea157fe565998a5f3e349", "messageHeadline": "Cannot upgrade SDK version because winml_lib_telemetry pulls in SDK c\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 13, "deletions": 3}}, {"node": {"oid": "13a129054f00562b08784d0a8d9ff318bd8b8063", "messageHeadline": "Prevent unnecessary re-initialization of the graph when model has unu\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 39, "deletions": 7}}, {"node": {"oid": "c4f515d38083548db94791f2bb383b50416672d4", "messageHeadline": "- Fix training cmake file so it builds if `--cmake_extra_defines onnx\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 153, "deletions": 150}}, {"node": {"oid": "de4c2217123718a8975007557e350af86c13ad82", "messageHeadline": "set max profile before opt profile (#7801)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "ae14cedd6336bc3114d66d26ca45fa48baab82ee", "messageHeadline": "Fix c_api warning (#7803)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "21ff8fabe933524e53fb5c2f22befa182671b61c", "messageHeadline": "[js/web] fix webpack config for onnxruntime-web (#7785)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 24, "deletions": 18}}, {"node": {"oid": "ee29330cabf6075f190776d35d70096dde39905f", "messageHeadline": "Delete unused file: Dockerfile.ubuntu_gpu (#7797)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 39}}, {"node": {"oid": "b5c5e8c1cadf078a1fa1f163589ae4f3aded19f5", "messageHeadline": "Update C++ API comment to resolve warning. (#7776)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "1fbc04d691142370af3ece415146ccf424ef361f", "messageHeadline": "Enable training ops in inference (#7783)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 87, "deletions": 9}}, {"node": {"oid": "b852b73e84ff1f2d4e0462e682aac026b492b2e0", "messageHeadline": "[js][doc] update some part of documents. (#7768)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 7, "deletions": 95}}, {"node": {"oid": "a6ca9f0a402636bb25ce47d5349ac6c489ad3888", "messageHeadline": "Use list comprehensions instead of list appends where possible (#7753)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 101, "deletions": 44}}, {"node": {"oid": "db0d608ff02b6a89922f778c5c7596c3f48e8f73", "messageHeadline": "Fix build errors on Dev machines after PR #7626 merge (#7781)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "7c4a5faef57b2255544f6ca0f5b34cfb89792513", "messageHeadline": "[wasm] enable DWARF format debug info for ORT WASM (#7777)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 15, "deletions": 13}}, {"node": {"oid": "2a028711571ba0086b3a264409a34d40056fc0b8", "messageHeadline": "Disable reuse for YieldOp's inputs (FW partial graph's output) (#7767)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 19, "deletions": 6}}, {"node": {"oid": "62a1c39f753e27c1366559bb32cae7d4cd5fa320", "messageHeadline": "Merged PR 6070939: Merge ORT master", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 43612, "deletions": 6247}}, {"node": {"oid": "3eace1ac4800c92c3169122014fab993c1acb415", "messageHeadline": "React native debug keystore. **BYPASS_SECRET_SCANNING**", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "503bd59dccbf9dbeab3a6519c4d611b7f8df6742", "messageHeadline": "Merge commit '31e6d3f85c21bb3fc0a9b194cf5a6e7380fd57d8' into user/jus\u2026", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 43612, "deletions": 6247}}, {"node": {"oid": "c2435d24ecbeededf1dc50187ab3bd11ad4a6994", "messageHeadline": "Clean up ROCm4.1 Dockerfile build directory (#7732)", "author": {"name": "Peng", "email": "peng.sun@amd.com", "user": {"login": "sunway513"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "c91602070d1fd3c159240b62ea247686f41c9f0a", "messageHeadline": "[js] update version of package \"onnxruntime-web\" and \"onnxruntime-rea\u2026", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "6c252a0bea807aa748b8cfc5dbaad52af3a9e11e", "messageHeadline": "Fix builds that use gcc 5  (#7765)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f6eb0f76ae5b6f6a5cedf22ee8b2994eb9503361", "messageHeadline": "to used cudnn7 to build onnxruntime-training wheel with Cuda 10.2 sup\u2026", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 199, "deletions": 24}}, {"node": {"oid": "c99aa3a3f3834adcbb888ce4b964f2695c524eae", "messageHeadline": "Ryanunderhill/cuda shared (#7626)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 5388, "deletions": 3710}}, {"node": {"oid": "31e6d3f85c21bb3fc0a9b194cf5a6e7380fd57d8", "messageHeadline": "Revert CUPTI profiling feature (#7763)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 51, "deletions": 246}}, {"node": {"oid": "e26c668a9b8e4a2f9ae76a221ba62a81c04f768a", "messageHeadline": "add google benchmark as direct dependency (#7762)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 15, "deletions": 1}}, {"node": {"oid": "38c1c6d2da14b4a9c4b4f4a01717bc758cf5df1d", "messageHeadline": "using copy to when scale/zp are same in qlinear concat. (#7761)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 61, "deletions": 23}}, {"node": {"oid": "374acf14238b73ebf9596800e3dfdf1377fb39fa", "messageHeadline": "Disable external initializers build option (#7635)", "author": {"name": "alonre24", "email": "alonreshef24@gmail.com", "user": {"login": "alonre24"}}, "additions": 130, "deletions": 13}}, {"node": {"oid": "6c868341e3398b6f81f3d5b098f9085cf1248914", "messageHeadline": "Fix CUDA 10.2 pipelines (#7759)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 73, "deletions": 16}}, {"node": {"oid": "fbe6eccc65228193365e01c1c705ec51cd9c4cb5", "messageHeadline": "[JS/Web] Bug fix for Reshape Pack (#7754)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "203efef1472859dc9aef65592a1f608ab9a1343f", "messageHeadline": "Add a flag to allow for linking against shared system protobuf (#7577)", "author": {"name": "Uwe L. Korn", "email": "xhochy@users.noreply.github.com", "user": {"login": "xhochy"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "3a68c389d97e3ef9f0df0f73ea3336bbe8b767d1", "messageHeadline": "Add version lock to manylinux build scripts (#7755)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "126afbe450440f79b8fd4c13d7b3432a34a1ee28", "messageHeadline": "Fix shape inference warning issue in TensorRT (#7751)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 26, "deletions": 26}}, {"node": {"oid": "47b3cc4bded14a633ee018881561f44e24b01b21", "messageHeadline": "GatherGrad Bugfix (#7752)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 43, "deletions": 4}}, {"node": {"oid": "75dbc2a35eae0c8caf67c14584f2cd9237b2b1ad", "messageHeadline": "use stable onnx.ir_version in test (#7747)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 18, "deletions": 18}}, {"node": {"oid": "077e8c6b40b648d69df112bf91f3f7cf7751f7d4", "messageHeadline": "allow update_version.py to update new npm packages (#7746)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "3f204d191b0e4fb71873e71d83cbeb3452a9e1ce", "messageHeadline": "[CoreML EP] Add Squeeze Op support  (#7730)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 137, "deletions": 18}}, {"node": {"oid": "d1c531058ae93a130a56325b89ec157f3c9766aa", "messageHeadline": "Add elseif statement for arm64e", "author": {"name": "Taewoo Kim", "email": "72404945+aud-ktw@users.noreply.github.com", "user": {"login": "aud-ktw"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "1e6ad669cf7a25a940a961f808a0a8da41b79d21", "messageHeadline": "Support arm64e for osx", "author": {"name": "Taewoo Kim", "email": "72404945+aud-ktw@users.noreply.github.com", "user": {"login": "aud-ktw"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e05b15175db366200deef19e2cba00b489949bd8", "messageHeadline": "Add cpp ext lock file check during ORTModule init (#7740)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 205, "deletions": 171}}, {"node": {"oid": "224a664811508e59f44e733490f46849cce8c2f1", "messageHeadline": "GPT-2 one step search tutorial (#7718)", "author": {"name": "Xiaoyu Liu", "email": "lixiaoyu@microsoft.com", "user": {"login": "xi-liu-ds"}}, "additions": 485, "deletions": 0}}, {"node": {"oid": "7834ca983c2fafd4705313a8aaa1cd5743724e4d", "messageHeadline": "update optimizers for opset14 (#7722)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 59, "deletions": 54}}, {"node": {"oid": "26a472c948a56ce46d9a27863fa4c124e0d3de9c", "messageHeadline": "Increase test timeout from 1 hour to 2 hours (#7735)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e161213f8efa3e9cfcd005dbc58f025199e810f7", "messageHeadline": "Handle model with no parameters (#7736)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 65, "deletions": 11}}, {"node": {"oid": "96deec596fb8bfca249723aae2f7ed6427c1ac93", "messageHeadline": "fix npm test for webgl without wasm artifacts (#7742)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "e92b3c13948904d8ba503690ce5e9ea1bb158d91", "messageHeadline": "bumping up version number to 1.8 (#7733)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 18, "deletions": 12}}, {"node": {"oid": "e4a985ff17a0be0b7b39050a9890f5fa1f28ee45", "messageHeadline": "[JS/Web] WebGL Profiling Tool (#7724)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 67, "deletions": 18}}, {"node": {"oid": "43e2ee37f22d6188fd666dfea48dcdf14f9f0287", "messageHeadline": "Some cosmetic changes (#7741)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 31, "deletions": 31}}, {"node": {"oid": "a6972c8782f1477b8ef557e2c54ac0075beacdb8", "messageHeadline": "Fix issues in TensorRT provider options (#7738)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 66, "deletions": 82}}, {"node": {"oid": "e9057d2e498e0c9233669cae11223bf4f6896ba6", "messageHeadline": "ZCode FastFormers changes (#5827)", "author": {"name": "Young Jin Kim", "email": "youki@microsoft.com", "user": {"login": "ykim362"}}, "additions": 112, "deletions": 22}}, {"node": {"oid": "38d90b0f1577ba18c896c700cf4292976da1a986", "messageHeadline": "Cleanup install_deps.sh (#7734)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 16}}, {"node": {"oid": "9075488368d03afa70c1fa25b40691fc3ebb9f8f", "messageHeadline": "Add per-column support for QLinearMatMul (#7729)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 172, "deletions": 60}}, {"node": {"oid": "d3c4b70ede113792a40aee2a7d6fdfcb86a2c40d", "messageHeadline": "[Web/JS] Fixing two bugs in reshape_pack and im2col_pack (#7689)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 65, "deletions": 34}}, {"node": {"oid": "79854dda8f6218b665f34092ea0b508961e9fb03", "messageHeadline": "bumping up verson (#7731)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "f97764432430cef27dfd6782338604ed1f0a0a14", "messageHeadline": "ROCM support int reductions", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 116, "deletions": 191}}, {"node": {"oid": "da5f24bd2d8997177f4ed29019894d623cd6a39e", "messageHeadline": "Support additional session options and run options in WebAssembly (#7\u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 426, "deletions": 230}}, {"node": {"oid": "6d9f541442323b42b19a548a0d0ff93b21e3aa06", "messageHeadline": "[JS]moved logging level flag to global env (#7700)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 119, "deletions": 52}}, {"node": {"oid": "5e8086ad8ed395f01f585ae6e30ae369ce2d23d0", "messageHeadline": "Support fusions inside subgraphs in optimizer tool (#7701)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 168, "deletions": 54}}, {"node": {"oid": "ad95b193226546084aac9da510e59065fe55ea15", "messageHeadline": "Help build pass on ubuntu 16.04 with old g++. (#7706)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0e48187b4e8b7e7f629e725283751e06c11103d3", "messageHeadline": "Add type checks for QDQ transformer (#7715)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 472, "deletions": 88}}, {"node": {"oid": "c873f5589df943a9451bec1a0d4ac5e92f2a8808", "messageHeadline": "Fix bug where the output names were sorted lexicographically (#7709)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 46, "deletions": 6}}, {"node": {"oid": "6c41ed597b0ffc003ce9e25315c4b7ac0f4850db", "messageHeadline": "Add custom autograd function to prevent input passthrough on ORTModul\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 139, "deletions": 21}}, {"node": {"oid": "4fe2ffae168c779a2e1267c3b42548eb40cde58f", "messageHeadline": "Fix ORTModule python doc generation (#7704)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 18, "deletions": 25}}, {"node": {"oid": "ebee3809110f8c32ce5f77886fa1f4b4edf7806f", "messageHeadline": "Partially revert PR #7348 (#7702)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 13, "deletions": 11}}, {"node": {"oid": "2b73163690654a8896e4ae62b2581afef9f5a200", "messageHeadline": "Exclude training specific logic from BatchNormalization to reduce bin\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 38, "deletions": 15}}, {"node": {"oid": "557b94637dfabea9db367c7a5ae56ed9f4880748", "messageHeadline": "Add more TensorRT env variables to provider options (#7698)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 422, "deletions": 112}}, {"node": {"oid": "943ab9dcefefcc89ca6ab6f78c7d3ee30f3ba7b1", "messageHeadline": "undo clang-format corruption (#7720)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 787, "deletions": 492}}, {"node": {"oid": "d604281a8604be362d57ef1483550a13272d7ed7", "messageHeadline": "Liqun/training pkg to run tests (#7662)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 89, "deletions": 4}}, {"node": {"oid": "3ead2f2f3965c0baa1eafa38c83e03c6847c21e5", "messageHeadline": "update pt lightning version (#7711)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "017b94ab366eafead3a94b10b8510dc372212ae7", "messageHeadline": "Specify correct dependency for CI pipeline of nodejs binding (#7717)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6b0a7905ede439304987d5d6cb3f1e4d147e5930", "messageHeadline": "fix quant weight cleanup bug (#7707)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 23, "deletions": 2}}, {"node": {"oid": "9ba8da65d2c62324382b2e6a29cc3a69b0e94278", "messageHeadline": "Fix BiasDropoutFusion when there are multiple Dropout consumers. (#7708)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 87, "deletions": 42}}, {"node": {"oid": "53d1d55ea88f9c8a7da74dd67608a75236db9e7b", "messageHeadline": "Add ability for pre-packed weights of shared initializers to be share\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 3466, "deletions": 1037}}, {"node": {"oid": "c78a40c0c79f1d6a80072096d4801b21fe379173", "messageHeadline": "add opset14 registrations for cuda provider (#7699)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 398, "deletions": 182}}, {"node": {"oid": "97d9bcd644e1c25c08e00a85a08211993c361f05", "messageHeadline": "[js/web] fix bundle for multi-thread, add e2e test and support nodejs\u2026", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 612, "deletions": 145}}, {"node": {"oid": "a74e41e47d5ccb02773dcf8de64b3de9a47210c2", "messageHeadline": "Add non-zero zp support for quant matmul and attention (#7570)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 719, "deletions": 335}}, {"node": {"oid": "c53b5be5093a60d0c18de009bc3b7bbb672d5a78", "messageHeadline": "force multi steps to use the same commit in CI (#7697)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 51, "deletions": 4}}, {"node": {"oid": "0f7721a0194ebd7c2908c0b3cecabd1be10d745c", "messageHeadline": "Fix bug for not checking original float value names when treat it as \u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "1d403ba03b32b195f7a84db10267d34f02d219b3", "messageHeadline": "Fix a compile warning in EigenNonBlockingThreadPool.h (#7638)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "033f0b3b7c4b6c1bf70747df3d7ca68c42c9a2d3", "messageHeadline": "fix typo. (#7690)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "a9b47ca8e7bc51974ec4401c93895b80f7c0080c", "messageHeadline": "update debug.keystore for a public usage (#7696)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "359fe1d197fecff4169e1446f4d5eb888cc1fbb8", "messageHeadline": "Liqun/ort training version (#7620)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 223, "deletions": 12}}, {"node": {"oid": "bfbcc89db1adc8689b1b9dde18e5d2872ea40c87", "messageHeadline": "Add MLFloat16 support for SoftmaxCrossEntropyLoss for CUDA EP (#7679)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 318, "deletions": 112}}, {"node": {"oid": "39fac6d30415ad14d9f7748e9d48b1b1f8592c43", "messageHeadline": "Fix a buffer reuse bug in allocation_planner.cc (#7645)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "442c7300eb30d57ded968480bdcaf55a6d853f72", "messageHeadline": "add opset14 rnn ops (#7687)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 74, "deletions": 32}}, {"node": {"oid": "3725d0211fb4ac26197de71f7f7f05443dbb58a0", "messageHeadline": "support maxpool QDQ fusion from opset 12 (#7693)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 33, "deletions": 11}}, {"node": {"oid": "50c5edcf13ac6db8e6bfcaf415fee87715db70ef", "messageHeadline": "Add nhwc support for QLinearAveragePool operator (#7656)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 741, "deletions": 197}}, {"node": {"oid": "a27ef39d8f0eff2738d1b8f68c379ca7659c0857", "messageHeadline": "Merged PR 6047693: DML EP fix GetSupportedDeviceDataTypeMask for int6\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 25, "deletions": 24}}, {"node": {"oid": "37f69fcee52d84aedded4be071da9c8d890408e3", "messageHeadline": "Regain performance by caching initializer names in ORTModule (#7685)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 81, "deletions": 37}}, {"node": {"oid": "19704aedbbd78f6ae53746c0968a7c22a839f5c3", "messageHeadline": "Update Objective-C API (#7675)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 477, "deletions": 82}}, {"node": {"oid": "56e993a434ff1f1cbce779d9e4a3d9b77e3b90f4", "messageHeadline": "Bump to rel-1.9.1 (#7684)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "32d8278c2fc1f93b73df53d266c309044ccc41dd", "messageHeadline": "reshape fix (#7678)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 11, "deletions": 9}}, {"node": {"oid": "4b37901f10555fb6ded76c63114b75a31a2f8e6b", "messageHeadline": "Aten support for rocm (#7680)", "author": {"name": "raviskolli", "email": "48601275+raviskolli@users.noreply.github.com", "user": {"login": "raviskolli"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "4afdc19958aa2814deb2af995c5e56b34f8989b9", "messageHeadline": "ROCm optimized layernorm for MI100 (#7682)", "author": {"name": "Aswin John Mathews", "email": "81309834+amathews-amd@users.noreply.github.com", "user": {"login": "amathews-amd"}}, "additions": 43, "deletions": 2}}, {"node": {"oid": "d90a99aad59f91d54d5c69ca5050084bc5b96928", "messageHeadline": "Fix the build on dev machines by replacing std::tuple with two argume\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "7bb3f243ff7de673e042ad3f4860539731a3ecb8", "messageHeadline": "Revert (#7663)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 12, "deletions": 393}}, {"node": {"oid": "1ab8a95eb6675afb6d0ad9d93600ef0022e2ddb5", "messageHeadline": "Bind existing SessionOptions and RunOptions in Javascript API with We\u2026", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 598, "deletions": 175}}, {"node": {"oid": "333318af04285da2e8279cba23a4d1e6f7491776", "messageHeadline": "[CoreML EP] add clip support (#7666)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 222, "deletions": 60}}, {"node": {"oid": "46246f1bbdb518454fed997087576da8367f220f", "messageHeadline": "Add nhwc transformer support and unittest for qlinear concat. (#7587)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 123, "deletions": 9}}, {"node": {"oid": "3a407b40ddf02bc549939ded748712600c6a4fa7", "messageHeadline": "Add the missing source file to the target onnxruntime_test_debug_node\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "31ca21b78268ff7c71f409fd53830a5e0cb210e1", "messageHeadline": "Replace Where Grad \"Mul\" with \"Where\" (#7672)", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 5, "deletions": 6}}, {"node": {"oid": "b4e8e9b004cb3a5b5f4b9cd59212fe7704e70538", "messageHeadline": "Add  DnnlOpManager (#7521)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 389, "deletions": 70}}, {"node": {"oid": "dac24f7d63f14c1d3f92f04ab6c7afeb65564819", "messageHeadline": "Add ATenOp and call aten::embedding and its Backward Op from ORT (#7590)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 723, "deletions": 47}}, {"node": {"oid": "c808621e9dd3f51f8cbc30b1d3edf4d20660bba2", "messageHeadline": "Better detection of MPI using FindMPI (#7653)", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 21, "deletions": 42}}, {"node": {"oid": "a47a234b7e3e858ffcbb2b701c06137fcfbf207e", "messageHeadline": "Add minsdkver for AAR and AndroidTest (#7669)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "760828b2d48d619736f577d714fc06fa7e79bd11", "messageHeadline": "Add FromProviderOptions()/ToProviderOptions() for TensorRT EP (#7654)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 159, "deletions": 15}}, {"node": {"oid": "1c7e683a95bb3c23f780ecbe8db68edf61ef037a", "messageHeadline": "Add Squeeze and Unsqueeze support for quantizaton tools. (#7673)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 121, "deletions": 0}}, {"node": {"oid": "31d4413919293f77d1b0f67077968112c9c7200a", "messageHeadline": "fix quantization tool bug when existing pass through only input (#7674)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "7cb9077043974e1863d08d76b381797246129594", "messageHeadline": "Fix readme page (#7659)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 22, "deletions": 317}}, {"node": {"oid": "9241f62e4c74b32f99db4d0fb2258520468cb207", "messageHeadline": "enable MatMulScale and cast propagation for ROCm EP. (#7657)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "5d9885f70602d1f9bcd58e1c61f00b9c1245704b", "messageHeadline": "Fix BadNames. (#7658)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 24, "deletions": 20}}, {"node": {"oid": "8deca24b1a66bd97b3593a3f02fab47a809e0ded", "messageHeadline": "Don't remove an unused initializer if it is overridable. (#7649)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 50, "deletions": 21}}, {"node": {"oid": "c5aeaa94196bbcbdead24f37dae930d744fc2186", "messageHeadline": "Support for unused model initializers (#7631)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 83, "deletions": 24}}, {"node": {"oid": "88d2fc8f1e3b6dea88f4431778cd7af7b325bec2", "messageHeadline": "ONNX Runtime React Native Library (#7564)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 20243, "deletions": 6}}, {"node": {"oid": "29172d8f54a645deb28fccec45edbe7758a774c5", "messageHeadline": "Setup EP Dashboard (#7321)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 257, "deletions": 95}}, {"node": {"oid": "ce8473a4ea690ae99b281d12d40028cc7c620361", "messageHeadline": "Add script to build fat iOS framework (#7607)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 170, "deletions": 0}}, {"node": {"oid": "d39db89fbb12894b76ca3e935abc54357bf42815", "messageHeadline": "Add info on some additional pytorch models that were added to the tes\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "5276bab268cffbc9fa93062a81b8a773ab56281f", "messageHeadline": "Fail model loading if node input is a missing value (#7459)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 50, "deletions": 3}}, {"node": {"oid": "90c65ac171046e19ef344d5a45661b9f9c6e5131", "messageHeadline": "Fix a bug in inference_session.cc (#7639)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "69d1db83ac3e8430e38350ddccca4bebe57e7865", "messageHeadline": "Enable bitcode for iOS by default (#7640)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "9f69b2f291d3633450305505f5ca48b0d556f1bd", "messageHeadline": "Added InsertAndReduce strategy to PropagateCastOps transformation in \u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 1437, "deletions": 487}}, {"node": {"oid": "08fbfe96074599e329254ea711174ff0415d0ddb", "messageHeadline": "Resolve issue where a registered buffer was parsed incorrectly as a u\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 55, "deletions": 10}}, {"node": {"oid": "a684e9aa529ff5f115653326290ede3079297d0f", "messageHeadline": "Add pre-training transform to convert BatchNorm to BatchNormInternal \u2026", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 301, "deletions": 51}}, {"node": {"oid": "de4089f8cbe0baffe56a363cc3a41595cc8f0809", "messageHeadline": "GCC11/Libstdc++11 Compilation fixes  (#7599)", "author": {"name": "ankurverma85", "email": "31362771+ankurverma85@users.noreply.github.com", "user": {"login": "ankurvdev"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "16297a8e612e8a31fa6df048c2d0fe33d92ca856", "messageHeadline": "Implement NCHWc Upsample linear mode (#7623)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 414, "deletions": 71}}, {"node": {"oid": "ec885040ef5f6d76dcc38a128a097f61d11df629", "messageHeadline": "[js] specify correct config for terser (#7627)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 305, "deletions": 480}}, {"node": {"oid": "c5d28097e8eb0291f8f7a3f3839309ff92479d66", "messageHeadline": "[js/web] adding conv fuse logic (#7604)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 91, "deletions": 16}}, {"node": {"oid": "88c95ef06bf81b829fb2e194131af01b34778d62", "messageHeadline": "Support for primitive types in ortmodule (#7588)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 119, "deletions": 10}}, {"node": {"oid": "4b691a5c0d9d48e2c899291926e47848b621b0e0", "messageHeadline": "Add ability for memory arenas to \"shrink\" periodically (#7284)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 785, "deletions": 245}}, {"node": {"oid": "803837df63ab9b36a6aacf7c5686fabaef19a6dc", "messageHeadline": "Add 4dmask support for attention cuda kernel  (#7591)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 395, "deletions": 27}}, {"node": {"oid": "55c086b6640f78a3b69d43c078b2288aa53d7920", "messageHeadline": "symbolic shape inference improvements for contrib ops (#7606)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 90, "deletions": 38}}, {"node": {"oid": "5a5fec0452740f93c6d0f7ac73ee455983a338a7", "messageHeadline": "Fix logs getting skipped in single-line conditionals. (#7589)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 199, "deletions": 105}}, {"node": {"oid": "e91bdbde2050d8a6db75037bfe7be6b5c4a50f8a", "messageHeadline": "Add myself to CODEOWNERS for ORTModule python code (#7453)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 4, "deletions": 4}}, {"node": {"oid": "41e370c2b30defa41174086375d391ff965d38c3", "messageHeadline": "Update protobuf to 3.16 (#7616)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "3c39fcc1fa0a71d8595b0a22ec2ac35143c78805", "messageHeadline": "[js/web] port fixes for packed concat over to ort repo (#7605)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 30, "deletions": 42}}, {"node": {"oid": "bdefc6c4d8c6f8daa60433139a4e12a8f5746fb3", "messageHeadline": "[js/web] support multi-thread for wasm backend (#7601)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 380, "deletions": 228}}, {"node": {"oid": "8ab0deceedc01aa3af5de60cdd2b2c0fd4b00650", "messageHeadline": "Add DLA support to TensorRT EP (#7532)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 67, "deletions": 17}}, {"node": {"oid": "9fc4116d51f909e1ac1e2db289c0a1c6fc7992d6", "messageHeadline": "Use ASSERT_STATUS_OK so the error message is output if there's a fail\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 10, "deletions": 20}}, {"node": {"oid": "0c91b643fed7f6be75ec2d7c75e9c4a26738e86f", "messageHeadline": "Bugfix for Scatter and GatherElementsGrad (#7593)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 54, "deletions": 5}}, {"node": {"oid": "cea0ea1591667799b073c58c24543ea5501fcefa", "messageHeadline": "[OpenVINO-EP] Remove support for 2020.4 (#7580)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 28, "deletions": 55}}, {"node": {"oid": "bdb2ed786411fb73c98b8164d422f88be13ff570", "messageHeadline": "Revert \"Add log to allow serving platforms to quantify ORT usage. (#7\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 25}}, {"node": {"oid": "d88da4406604d7468de9e879f2620dcdc72dd30f", "messageHeadline": "Allow flexible order of Add inputs in Attention fusion (#7565)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 53, "deletions": 25}}, {"node": {"oid": "48967446389eab23c0a834fb3e4129c94c9d0dab", "messageHeadline": "Improve CPU node placement logic to avoid Memcpys (#7427)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 393, "deletions": 12}}, {"node": {"oid": "097bab8d1e1230e35b85bb169618390c3a94ca98", "messageHeadline": "Cleanup a change to ExecutionFrame a little (#7576)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 44, "deletions": 73}}, {"node": {"oid": "be2a3046fe67039cd1a83c9eee98f37c98fcafb5", "messageHeadline": "Disable telemetry log for training builds (#7585)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5413eaa5e4ab3283e6ae6e827b4e5593a0120a04", "messageHeadline": "Additional cmake changes for OpenVINO build (#7579)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "830f0b45d01f8472518794147759ddffb9b9c2db", "messageHeadline": "Update Objective-C API (#7567)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 355, "deletions": 207}}, {"node": {"oid": "f3a70f1aec278ac287fac6e82dbd50bd78891a5c", "messageHeadline": "Ignore invalid input argument to install_os_deps.sh (#7566)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "94659487159004ab0edb7a35a1d03a2384bac66c", "messageHeadline": "Quantization tools using one more extra_options on interface. (#7293)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 40, "deletions": 20}}, {"node": {"oid": "a284eede64d206aca27d05a98c6f38a1209ff33c", "messageHeadline": "Fix Linux CPU pipeline (#7584)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 5}}, {"node": {"oid": "8a9ddfe963cc2eed511eb2f351f1e9b4b5e23aab", "messageHeadline": "Longformer Attention non-determinism issue fix (#7574)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 13, "deletions": 6}}, {"node": {"oid": "94c97ac8c2b2d9459e054a304cdc985ea9499dfa", "messageHeadline": "Fix compiler warnings treated as errors in GistEncodeDecode. (#7568)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 11, "deletions": 10}}, {"node": {"oid": "91985ab03d310c5a7af0f7c96bb93299777bf587", "messageHeadline": "add use_dml (#7569)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 6}}, {"node": {"oid": "2f0479780e4abfa78deb35ad703b832e84bc4033", "messageHeadline": "Improves NonMaxSuppression on CPU (#7557)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 43, "deletions": 58}}, {"node": {"oid": "ade6ed51eb81415f9273c77b2eaa7b47ef54c677", "messageHeadline": "Speed up Reduce operators for consecutive reduced axes (#7206)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 2277, "deletions": 249}}, {"node": {"oid": "053bada30fe6d027063a9ca10efbec08edf69776", "messageHeadline": "Add support for setting shape inference function on fused nodes (#7007)", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 86, "deletions": 1}}, {"node": {"oid": "d8cf9604123e078b2f12e7c13902ba3a0ae56244", "messageHeadline": "Add android test app to validate Java API for ORT-Mobile Android (#7477)", "author": {"name": "Rachel Guo", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 571, "deletions": 14}}, {"node": {"oid": "f6cefc92e29c6800df440b88670b92a54419bf7e", "messageHeadline": "Add quantized value map after quantize input node added. (#7558)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 20, "deletions": 29}}, {"node": {"oid": "a647da3e1acb5ecb750c19df6b1b85af4572237d", "messageHeadline": "Fix 2 input Gemm grad (#7561)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 11, "deletions": 3}}, {"node": {"oid": "d812354ebdbc13848f846889d8bf32ac41731c3b", "messageHeadline": "Tile grad fix  (#7556)", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 3, "deletions": 8}}, {"node": {"oid": "e05528a365c7672ca76a453f679bc449a84466ca", "messageHeadline": "Update Android AAR packaging pipeline script (#7559)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 11, "deletions": 7}}, {"node": {"oid": "71ff6ff2eceba221be68b6911dca936ce204e53d", "messageHeadline": "Disable NNAPI support for dynamic input shape, add warning logs (#7439)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 66, "deletions": 20}}, {"node": {"oid": "c3c4db2c1b67cc9976fbdbb61e0c1708d3b9cbb0", "messageHeadline": "Upgrade GIST memory compression nodes, kernels, optimizer rule, and c\u2026", "author": {"name": "Fanny Nina Paravecino", "email": "fninaparavecino@gmail.com", "user": {"login": "fninaparavecino"}}, "additions": 1669, "deletions": 116}}, {"node": {"oid": "9456bf420ea6570c9f01d0510f3a42fa673a5707", "messageHeadline": "Merged PR 6006709: Merge onnxruntime's master with DmlDev", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 17890, "deletions": 4788}}, {"node": {"oid": "c1ed64717013b5234d928a8724e901046e78963b", "messageHeadline": "ORTModule enable run_symbolic_shape_infer by default (#7423)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 15, "deletions": 7}}, {"node": {"oid": "a94a893d5ebd452ef164fa329dc19248d2d4fb86", "messageHeadline": "Update SessionOptions.cs (#7540)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "729027078874169466934140b22530a65c3ec2ae", "messageHeadline": "Merge remote-tracking branch 'mainstream/master' into DmlDev", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 17890, "deletions": 4788}}, {"node": {"oid": "594dde264723c3f57ff50d6088e335332bcb8a22", "messageHeadline": "Validate that the conversion script from the python package can be us\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "898fff702ce2b4ba9a8b13465a7b5cc25d664709", "messageHeadline": "compatibility was broken for myriad config parameter (#7349)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 21, "deletions": 16}}, {"node": {"oid": "3c9ece4a1106a4e206fdf825c55cd0c9c0608586", "messageHeadline": "[transformers optimizer] catch symbolic shape inference exception and\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 51, "deletions": 125}}, {"node": {"oid": "faea7a222da3ac67efafd030e5942cf8d23ebdf5", "messageHeadline": "linux trt package pipeline (#7537)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 189, "deletions": 1}}, {"node": {"oid": "8eaa4c33e206991ba09a1d7d5138fee416c1df33", "messageHeadline": "[js] fix library bundling and some trivial improvement (#7550)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 135, "deletions": 106}}, {"node": {"oid": "731f9e5033eb9f86125d5768a3dbb563696b7147", "messageHeadline": "Fix symbolic shape inference for Unsqueeze (#7555)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 119, "deletions": 41}}, {"node": {"oid": "418623355ae263664e7440a72dbc3b72dcd24caf", "messageHeadline": "disable logging for WASM in inference session ctor (#7545)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8b6602ae687925a8125029cdd123a4ec8c235ca7", "messageHeadline": "Refactor provider test utils to prepare for expansion. (#7538)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 262, "deletions": 225}}, {"node": {"oid": "cab84d902eccf9de53ada9e0c6473aecc1e8246c", "messageHeadline": "Install and use conda on ortmodule CI pipelines (#7530)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 140, "deletions": 79}}, {"node": {"oid": "ad15811adece27b3dd4f4f132afc5481b8f8f275", "messageHeadline": "Add QDQ support for MatMulIntegerToFloat, Gather and Transpose (#7500)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 312, "deletions": 90}}, {"node": {"oid": "830d9e54dddbc92c3cd25087d7b087752fe96bbf", "messageHeadline": "Add script to dump initializer, NodeArg, Node and subgraph info from \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 145, "deletions": 0}}, {"node": {"oid": "3600c3e66e0b068d9d1e5d71b7d6e27106a00813", "messageHeadline": "[js/web] integrate latest changes from onnxjs (#7535)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 9440, "deletions": 503}}, {"node": {"oid": "9c0e5954cb86ac7e471de3901fdf8465b4540604", "messageHeadline": "Output Tensor Shape Validation b/w ONNX inference and ORT  (#7252)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 140, "deletions": 7}}, {"node": {"oid": "e344a583b0f39c6b8aca57f61b957022f45ec047", "messageHeadline": "updated sampleTolerance of model fp16_inception_v1 for GPU execution \u2026", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "b0a3b501fe13aaba6b8fce435469ca36adfe62d8", "messageHeadline": "Add function body for Gelu and FastGelu (#7496)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 120, "deletions": 4}}, {"node": {"oid": "7079dfb93dc16119b1ec030525b2e4a401942daf", "messageHeadline": "[wasm] fix and unify webassembly target name (#7549)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 45, "deletions": 27}}, {"node": {"oid": "2e09d9921a16faa82169eb9002cb0eb6c69b1461", "messageHeadline": "\"Sticky\" allocation of worker threads (#7551)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 552, "deletions": 373}}, {"node": {"oid": "6714f2f85d6af8912dd0027d6364b988ad41278e", "messageHeadline": "Improve tol value logging in ORTModule test (#7544)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 55, "deletions": 45}}, {"node": {"oid": "d1cb8c9dc9807fe79bf02921f719081c417f8aad", "messageHeadline": "Support negative indices and fix bound checking in symbolic shape inf\u2026", "author": {"name": "Ryota Tomioka", "email": "ryoto@microsoft.com", "user": {"login": "ryotatomioka"}}, "additions": 133, "deletions": 12}}, {"node": {"oid": "8e3cdf0452f333fa1fe13a84770d1c9c7591b097", "messageHeadline": "Use unicode apis for loadlibrary (#7523)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "add4e4225ba69ba48a28889ff91e65bbc5f6f2ca", "messageHeadline": "[js/web] fix pacakge metadata of onnxruntime-web (#7543)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 38, "deletions": 4}}, {"node": {"oid": "97de078c247158f4053711c641b9c0a08effab61", "messageHeadline": "[js/node] fix pacakge metadata of onnxruntime-node (#7542)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 19, "deletions": 23}}, {"node": {"oid": "79dc7d3e50dce9be46780c64d414e6107c30693c", "messageHeadline": "[js/common] revise TSDoc of some interfaces (#7541)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 64, "deletions": 59}}, {"node": {"oid": "8ba6ed953f47a918628ba4c0fb331aebce6cebb9", "messageHeadline": "Fix batch norm training op on CPU (#6946)", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 156, "deletions": 32}}, {"node": {"oid": "94c4c44bfc79257253ee3bc1db2f4ffd83c1cdac", "messageHeadline": "Enable Microsoft.Ai.MachineLearning package to work on .NET5 down to \u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "9f1e61be921a5e2e94d89e71f36eca893b457f03", "messageHeadline": "Check whether nvcc supports -Wstrict-aliasing before adding the flag.\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 31, "deletions": 9}}, {"node": {"oid": "00882ce4959634eae3f8d083985436233eb72878", "messageHeadline": "Set CMAKE_CUDA_STANDARD to 14 because we are using std::make_unique (\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 153}}, {"node": {"oid": "93e93d085137aeee5035da6cd319ebc2c3427547", "messageHeadline": "Merge pull request #7519 from microsoft/user/dwayner/ort1.8dml1.5.1", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 258, "deletions": 55}}, {"node": {"oid": "668a65f1a7150b7c88e1c530a3a8f1fceddf89ed", "messageHeadline": "Complete GetGlobalAveragePoolGradient  (#7514)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 46, "deletions": 30}}, {"node": {"oid": "9c1900866a599eb2eeba1a907ee5e5a22063c967", "messageHeadline": "Revert \"\"Sticky\" allocation of worker threads (#7372)\"", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 358, "deletions": 529}}, {"node": {"oid": "9ba9da0c956265a245b0753e6ca8804bd2fe7e8c", "messageHeadline": "Fix unused registered buffers issue on ORTModule (#7525)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 38, "deletions": 5}}, {"node": {"oid": "54db6648afbeecf1208dbb6c3308a1106f3e7826", "messageHeadline": "kerne invoker api for eager mode (#7473)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 342, "deletions": 7}}, {"node": {"oid": "dfca1a09d54dd01591270a33baf5467726041933", "messageHeadline": "Add Thread Spinning Session Option in WinML (#7498)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 171, "deletions": 6}}, {"node": {"oid": "e6f66f660cf2ad8bb25fc0e0d0188e664cb9b241", "messageHeadline": "missed change for external allocator in ROCm EP. (#7505)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "10e67b7340b9cc5f03feed5c9a0453ed1c4af967", "messageHeadline": "Merged PR 5918130: Add activation fusions missing in newer opsets", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 57, "deletions": 0}}, {"node": {"oid": "06a2b0401a9b6faa53ff94ce7d71259b8285b466", "messageHeadline": "Merged PR 5873494: Resize support nearest_mode floor in DML EP", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 73, "deletions": 20}}, {"node": {"oid": "e6f35cc132718a01dfe2337729044c276b1a7715", "messageHeadline": "Merged PR 5866812: Decompose unsupported QLinearSigmoid operation in \u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 118, "deletions": 8}}, {"node": {"oid": "f87527c0dfaec649cd4f4c4ee8de9a57e986770a", "messageHeadline": "Merged PR 5861108: Allow nodes in DML graph partitions with empty sha\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "915931384a1f78357b0408791ebcdafdb6da7f6c", "messageHeadline": "Merged PR 5807585: Remove support for strided 64-bit emulation in DML\u2026", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 3, "deletions": 21}}, {"node": {"oid": "70e67ddd2b86aa4d85a69d4b242d29ab54556459", "messageHeadline": "Update DirectML version to 1.5.1 and enable ARM/ARM64 builds with DML\u2026", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 102, "deletions": 32}}, {"node": {"oid": "00aaa6dabb7847ef50ac4e4344cd8a35261e77dc", "messageHeadline": "update CI for onnxruntime-web (#7497)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 192, "deletions": 42}}, {"node": {"oid": "0d107bbb738ce1f8ba9f94ca9b2d68041007919d", "messageHeadline": "Fix CUDA 10.2 pipeline (#7508)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d6df5764d72cc2ac3b3fd1fc6c87db9228109654", "messageHeadline": "Android package infrastructure (#7430)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 379, "deletions": 209}}, {"node": {"oid": "3d92723d1c9a78fa6af5e73e26c8f5393c156063", "messageHeadline": "\"Sticky\" allocation of worker threads (#7372)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 529, "deletions": 358}}, {"node": {"oid": "ec04b6203b632b83359cd87186dc6deee9a0a8e1", "messageHeadline": "Remove conditional compilation of std::is_trivially_copyable since we\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 20}}, {"node": {"oid": "1012535dabedd75e8b29e567e3edcaac64067752", "messageHeadline": "Change onnxruntime::make_unique to std::make_unique (#7502)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1070, "deletions": 1086}}, {"node": {"oid": "d337fa90e74b0985bc2fa3c2d9b8cf1c27e07b3f", "messageHeadline": "Propagate QDQ only when scale and zp are scalar (#7492)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 201, "deletions": 53}}, {"node": {"oid": "e255506bcd495116b23fd3f8f74ca7499d48be9d", "messageHeadline": "Add another input validation to ReverseSequence (#7445)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 37, "deletions": 1}}, {"node": {"oid": "994c2ed420f4764352404f5ccd91738f61bc1182", "messageHeadline": "GPT2 one step beam search update with configuration support (#7425)", "author": {"name": "Xiaoyu Liu", "email": "lixiaoyu@microsoft.com", "user": {"login": "xi-liu-ds"}}, "additions": 681, "deletions": 388}}, {"node": {"oid": "6358e96b63d4173535f400afe48384ff2254b3c5", "messageHeadline": "Added OpenVINO 2021.4 support (#7470)", "author": {"name": "Ilya Lavrenov", "email": "ilya.lavrenov@intel.com", "user": {"login": "ilya-lavrenov"}}, "additions": 81, "deletions": 35}}, {"node": {"oid": "7b003967b1c89b49a2fbda86f00efa0b5ea7e92e", "messageHeadline": "Add static code analyzer to Windows CPU/GPU CI builds and fix the war\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 521, "deletions": 486}}, {"node": {"oid": "2b0bbfd1a8524d0f684f606d4aa777e4a0fb3784", "messageHeadline": "MLAS: add SSE 4.1 u8s8 kernel (#7490)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 463, "deletions": 29}}, {"node": {"oid": "e73c3e06510acd53a63a81e160fe53482556ce85", "messageHeadline": "rollback the GetRuntimePath impl for linux (#7488)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 10, "deletions": 16}}, {"node": {"oid": "0dbe51b002d971168d232627cbc0735d611f25cd", "messageHeadline": "Enable TRT EP for C# (#7482)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 80, "deletions": 0}}, {"node": {"oid": "3c7c72898908a764cdccfc926ceed31625005c4e", "messageHeadline": "cmake: Add regex pattern for POWER architecture (#7494)", "author": {"name": "RajalakshmiSR", "email": "58744256+RajalakshmiSR@users.noreply.github.com", "user": {"login": "RajalakshmiSR"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f13b37899574c30192c5a4fc2add4766d418f2d2", "messageHeadline": "Re-disable tests (#7495)", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "e6a3308db7c03a13e0f08b221b6770e17fc3a4ef", "messageHeadline": "Optimize cuComputeGradInput performance. (#7479)", "author": {"name": "sabreshao", "email": "sabre.shao@amd.com", "user": {"login": "sabreshao"}}, "additions": 40, "deletions": 14}}, {"node": {"oid": "6773b4f5dd4c0a3a31e2d683883dafd3d3e2cb14", "messageHeadline": "Fix implicit-exception-spec-mismatch warning. (#7481) (#7483)", "author": {"name": "Chandru Ramakrishnan", "email": "41447659+chandru-r@users.noreply.github.com", "user": {"login": "chandru-r"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3ee63beafacbffe37d6a42abd7dc8d22d0c40d00", "messageHeadline": "Fix user input order before ORTModule feed it to backend (#7456)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 149, "deletions": 14}}, {"node": {"oid": "d68cedfa85c31efd10a0c13d9ebc579514a74d53", "messageHeadline": "Fix some C/C++ warnings in the jni part (#7385)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 47, "deletions": 23}}, {"node": {"oid": "ab373d6f0337fb82e1ae72a8af5b34dc3bd87048", "messageHeadline": "Lifhuan/force trt sequential (#7440)", "author": {"name": "Lifu Huang", "email": "lifu.hlf@gmail.com", "user": {"login": "lifuhuang"}}, "additions": 72, "deletions": 26}}, {"node": {"oid": "c584d482830b794dbb3f59dff8828546c4c6aca7", "messageHeadline": "Add sequence identity for opset 14 & fix sequence insert (#7335)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 103, "deletions": 39}}, {"node": {"oid": "22d7cde725e75de734c87529c7eb4eb4358d1fb4", "messageHeadline": "Fix a 'Squeeze' related issue in symbolic_shape_infer.py (#7380)", "author": {"name": "thilow", "email": "thilow@microsoft.com", "user": {"login": "thilow"}}, "additions": 38, "deletions": 1}}, {"node": {"oid": "674915208ae66134a58432c2ce28b8d9976af228", "messageHeadline": "Fixes RelWithDebInfo build issue on windows for OV-EP (#7471)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "044c78f089d97360d93d798367084f823b40758d", "messageHeadline": "Add function body to LayerNorm (#7378)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 480, "deletions": 272}}, {"node": {"oid": "da5c9263e93d7361da868970c51c334ab9cac278", "messageHeadline": "Add log to allow serving platforms to quantify ORT usage. (#7476)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "8e21329206f056dba82bc818941b449f526e47e2", "messageHeadline": "Update nuphar notebook model download url (#7475)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 9, "deletions": 9}}, {"node": {"oid": "196e6702ad004d20c6661885d3857d583268ca9f", "messageHeadline": "to support multiple cuda versions in published onnxruntime-training p\u2026", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 153, "deletions": 32}}, {"node": {"oid": "e64e30ee0d584bae66d77958326b35cd4de66805", "messageHeadline": "Improve ConvTranspose by transposing const filter during prepacking. \u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 260, "deletions": 70}}, {"node": {"oid": "9921f94ff66d70c627bae7c3d7b1a66e475b455e", "messageHeadline": "Merged PR 5982350: Latest Github master into Dmldev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 148682, "deletions": 13157}}, {"node": {"oid": "988ea0518e3d0f64f8acb134191d9b88a9e58b4b", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/rylai/latest\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 148682, "deletions": 13157}}, {"node": {"oid": "d21304ceb09d1b91160390ec80379b04c2bddedb", "messageHeadline": "Initial Objective-C API (#7366)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1131, "deletions": 33}}, {"node": {"oid": "78e583d08cf7a739e544a227a00c0ed3aec722f5", "messageHeadline": "Add CMAKE_CUDA_ARCHITECTURES=52 to TensorRT CI pipelines (#7455)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "c2418a1f42aa476000b41fc0ca42c7a96d6b6d17", "messageHeadline": "[wasm] fix memory info creation (#7461)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4cbd2cce9b41e888338a6f018ff873e85210f7ea", "messageHeadline": ". (#7466)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "4ebc9c3b5e4cbdd4cdf3e08906fef7ea49fed508", "messageHeadline": "[JS] onnxruntime-web (#7394)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 129412, "deletions": 43}}, {"node": {"oid": "d13e5b2fd908bc33c17eadde00a76ec3a5dc5408", "messageHeadline": "NCHWc: ReorderInput improvements (#7442)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 506, "deletions": 109}}, {"node": {"oid": "82108b18e34ee796c3c93ca146e048359f293fe0", "messageHeadline": "Partial graph execution perf improvements. (#7438)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 19, "deletions": 16}}, {"node": {"oid": "0702a14ee784e84ae7e9169508b54d22a7a3d45f", "messageHeadline": "Add pytorch version check before loading Python ONNX Runtime training\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 273, "deletions": 184}}, {"node": {"oid": "4804ede5018d1edac0a64d0c5967bbe16c3823dd", "messageHeadline": "Update build docker image cache cleanup build definition (#7452)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 14, "deletions": 2}}, {"node": {"oid": "40568d88212fe3c2072c621a834a0be932377e6a", "messageHeadline": "Wait for dispatch done in RunParallelSection to fix random TP UT cras\u2026", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 14, "deletions": 4}}, {"node": {"oid": "ada0fbbd2dc47cb9c4b08dd661500380525c7445", "messageHeadline": "Implement qlinear concat and unit test. (#7341)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1038, "deletions": 356}}, {"node": {"oid": "b5592856a7cd0be2242dea76825dcb34cc2a3f26", "messageHeadline": "Remove thread pool's cancel method and suppress some warnings (#7411)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 93, "deletions": 118}}, {"node": {"oid": "368e4a324f5235823cd8a948c505536b13e31ada", "messageHeadline": "SqueezeGrad Bugfix (#7412)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 24, "deletions": 22}}, {"node": {"oid": "ca9b3f18e9cb5a2ce686d5b13bfc76164c2a5c2a", "messageHeadline": "Explicitly pass cuda stream to thrust function rather than use cuda d\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 3, "deletions": 6}}, {"node": {"oid": "b9cbbc41ff786dbb781d01df196c00303d88f10c", "messageHeadline": "dnnl matmul tensor dimension check (#7383)", "author": {"name": "jeyblu", "email": "kiefer.kuah@intel.com", "user": {"login": "jeyblu"}}, "additions": 23, "deletions": 19}}, {"node": {"oid": "afe912d47c1214b1567846ab41326f78bb28ac87", "messageHeadline": "Reduce perf gap between thread pool and omp (#7333)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 94, "deletions": 91}}, {"node": {"oid": "410a81b21be2731c519f4023a190f48a0dc29538", "messageHeadline": "Add support for ORTModule to execute the graph when ONNX drops unused\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 120, "deletions": 31}}, {"node": {"oid": "f4f2cc1a00de102f289b1d3934948b3d6baba7a2", "messageHeadline": "Add batch interface to floating point GEMM (#7323)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 628, "deletions": 627}}, {"node": {"oid": "7a3c1787af012916fbabd109ffdf389dcde071d4", "messageHeadline": "Add CI pipeline to publish Python training package targeting Rocm (#7\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 359, "deletions": 7}}, {"node": {"oid": "34ebf7d3dd32e14295a409db08a667d6d236aeb6", "messageHeadline": "Partial graph execution made simple. (#7324)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 1081, "deletions": 502}}, {"node": {"oid": "5208231126a290fdddbed0714d6a23f3db46dd64", "messageHeadline": "Fix some warnings in our CUDA code (#7436)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "8889e717eb7f36b386c3aadadf47c318658f63d7", "messageHeadline": "add gather elements (#7435)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "ef72764960a90e9c33948b125d1a0937dbf90d7c", "messageHeadline": "Build would fail when nccl is not under standard path (--nccl_home) (\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 20, "deletions": 0}}, {"node": {"oid": "9f683bae7878f447a61912e8b3b45b6fd539dbc5", "messageHeadline": "Revert the TRT change and move the build to a new pool (#7434)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 39, "deletions": 13}}, {"node": {"oid": "979d63159bd9bbc98b0a7a2fd8738112731bc28c", "messageHeadline": "Add level two optimizations for constant propagation transformation. \u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 292, "deletions": 147}}, {"node": {"oid": "f1c3f3fcc1f7b0ef2be9291ebe1c5e3f8389bf7e", "messageHeadline": "TRT EP memory leak fix (#7415)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "043883b52def3a3b4bba0d22f3cabbfff7f254de", "messageHeadline": "[CoreML EP] Add Gemm/MatMul support (#7403)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 361, "deletions": 112}}, {"node": {"oid": "e7912736b9bfe82c003ad758e39068643d96373a", "messageHeadline": "Add qdq propagation support (#7404)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 844, "deletions": 93}}, {"node": {"oid": "1fa6d8fe1c1e41fcec15d63a57b3a0b5458b697b", "messageHeadline": "support loading external execution provider from python frontend (#7332)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 427, "deletions": 28}}, {"node": {"oid": "75e054cd3390747fcdbbf4e2dbf9964633cd926f", "messageHeadline": "pick onnx release candidate (#7177)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 338, "deletions": 213}}, {"node": {"oid": "d4140391899efffe4e593d93376d79417f6cc4c8", "messageHeadline": "Add ios coreml ci, and speedup ios ci run (#7420)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 24, "deletions": 5}}, {"node": {"oid": "d67c86265b68709cca9ffb5c6d241ee546ecd0d0", "messageHeadline": "Enabled fp16-inception-v1 test (#7406)", "author": {"name": "sumitsays", "email": "sumitagarwal330@gmail.com", "user": {"login": "sumitsays"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "b56dd037d3b22bf0125020d67e4f737657518099", "messageHeadline": "increase timeout for nodejs binding test (#7422)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4c8513a627c9b5a3fbcc172933e4f99c4ef1ed75", "messageHeadline": "SimplifiedLayerNormalization kernel for ROCM EP (#7409)", "author": {"name": "raviskolli", "email": "48601275+raviskolli@users.noreply.github.com", "user": {"login": "raviskolli"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "6822ae95eca01d70be67d906dd927026e75687d7", "messageHeadline": "Reduce the number of TensorRT tests needed to run (#7419)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 13, "deletions": 44}}, {"node": {"oid": "771a6d235b8495d05bcf6a906107df1bd6e81744", "messageHeadline": "Fix IsContiguousTensor check on backend (#7391)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 24, "deletions": 5}}, {"node": {"oid": "afa7b23609af8512dc1b98bc625d4ed58f03a1f2", "messageHeadline": "Update docs/ContribOperators.md and the script that generates it. (#7\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 60, "deletions": 19}}, {"node": {"oid": "1bbe538379861c02f5f4de62cb4a192e8aafa7e5", "messageHeadline": "Update references", "author": {"name": "Brian Popow", "email": "popow@gmx.de", "user": {"login": "brianpopow"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "aa1ce726aad49c8a84f931c46b846aaa36f809f2", "messageHeadline": "Remove unnecessary encoding step", "author": {"name": "Brian Popow", "email": "popow@gmx.de", "user": {"login": "brianpopow"}}, "additions": 3, "deletions": 9}}, {"node": {"oid": "65b2b87f838c35f23cdd902e43b1a2aa0e62d698", "messageHeadline": "Update CI build docker images (#7386)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 60}}, {"node": {"oid": "09313d9e1f2cb526330369eee88b489b23dd2c4f", "messageHeadline": "Added GreaterOrEqual and LessOrEqual Ops to RocmEP (#7398)", "author": {"name": "raviskolli", "email": "48601275+raviskolli@users.noreply.github.com", "user": {"login": "raviskolli"}}, "additions": 28, "deletions": 0}}, {"node": {"oid": "b4cfa88bf71441c4decda02aeee08c86f72d1c9e", "messageHeadline": "Update protobuf to the latest version (#7396)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 30, "deletions": 20}}, {"node": {"oid": "243713c4642b17c65e7561a16b6454601e0b8e38", "messageHeadline": "Upload detailed code coverage result to azure blob storage (#7392)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 34, "deletions": 3}}, {"node": {"oid": "16ca7677e6aa4c4eed1fe0fc1761bae55e212cc4", "messageHeadline": "Relax ConvGrad Test tol (#7393)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 14, "deletions": 34}}, {"node": {"oid": "b5493d724c42202b1a21a714c2558b0204dced41", "messageHeadline": "Update rnn_helpers.cc: add #ifdef to DumpMatrixImpl  (#7389)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "7b11283af03d0cb885470b98c1986be9cc11dcd6", "messageHeadline": "Add ability to allocate initialized tensor memory from non-arena memo\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 223, "deletions": 17}}, {"node": {"oid": "84211243447db1c0a36073b220fc72330ca2042d", "messageHeadline": "Add support to **kwargs in ORTModule forward() method (#7360)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 250, "deletions": 132}}, {"node": {"oid": "76cc118dbedd398f89a04cc1f625327827ae8132", "messageHeadline": "Gemm transpose fusion (#7306)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 333, "deletions": 0}}, {"node": {"oid": "913ea8264bbe2a3d942ce44690b3ed2675cba16c", "messageHeadline": "GPT2 with one step beam search (#7163)", "author": {"name": "Xiaoyu Liu", "email": "lixiaoyu@microsoft.com", "user": {"login": "xi-liu-ds"}}, "additions": 1291, "deletions": 8}}, {"node": {"oid": "1a3ddf0714e1bdf9b807a342eee5f6e160ad1ec9", "messageHeadline": "Add gradient registration and tests for Min/Max (#7217)", "author": {"name": "mindest", "email": "30493312+mindest@users.noreply.github.com", "user": {"login": "mindest"}}, "additions": 100, "deletions": 1}}, {"node": {"oid": "ce7ff27bac622aedb284ff3b57832055d715ae08", "messageHeadline": "Fix perf issue in  Conv CUDA kernel (#7348)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 83, "deletions": 30}}, {"node": {"oid": "ac346a1b90df59462779af10224a2ec9be4af790", "messageHeadline": "Modify SimplifiedLayerNormFusion to allow fusion in the presence of C\u2026", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 220, "deletions": 26}}, {"node": {"oid": "7abe1fd3920686a01d490a6530f2437fb2f221ad", "messageHeadline": "Identity elimination with graph output (#7312)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 265, "deletions": 4}}, {"node": {"oid": "265db2ad96b01c0ac32439214f482fb05706cc69", "messageHeadline": "Fix Microsoft.AI.MachineLearning .NET5 publishing and C# Store Releas\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "bb1e417da0abb79e711ae49737b8fc8642b5b753", "messageHeadline": "Add logging support to Cast Propagation transformation from python (#\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 221, "deletions": 57}}, {"node": {"oid": "e9286aa9a959ce956ababb33f999269080cfdddc", "messageHeadline": "Merged PR 5950370: Unsquash merge from github Onnxruntime into dmldev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "fc88a7aad532140204a08c2ec457112d1870e709", "messageHeadline": "Merge commit 'ef1aaa367aab7c0c384081678ae506e658be864f' into user/ryl\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "6dda1e068177aa51a85f3cde5a3079110451952c", "messageHeadline": "Flag for tensor memory re-use in allocation planner. (#7359)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 38, "deletions": 7}}, {"node": {"oid": "96cdc65d5775eab7cf12da9f592431b0d4f03c5a", "messageHeadline": "Fix android CI failure after gradle updated to 7.0 (#7364)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 22, "deletions": 0}}, {"node": {"oid": "009f342caf425edd1fe00832047d78d536d36402", "messageHeadline": "[JS] refactor Javascript/Typescript libraries in ONNX Runtime (#7308)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 8261, "deletions": 10054}}, {"node": {"oid": "ded2b0838060869244bc12ead2315df618cd8ab0", "messageHeadline": "WebAssembly multi-threads support. (#7326)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 99, "deletions": 58}}, {"node": {"oid": "28e229ac4c94805e7e84940ebd1e2d0435d033c5", "messageHeadline": "Enable build dynamic framework for macOS/iOS (#7343)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 104, "deletions": 16}}, {"node": {"oid": "e35922012560121fb2fde8de927ce169b817dbc6", "messageHeadline": "Merged PR 5938615: RI onnxruntime into Dmldev", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 37283, "deletions": 19572}}, {"node": {"oid": "ef1aaa367aab7c0c384081678ae506e658be864f", "messageHeadline": "Adding interface for batched integer gemm (#7249)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 783, "deletions": 440}}, {"node": {"oid": "f1c1c38d44d2545d598e3ea1739755fbdc1d4ad5", "messageHeadline": "Delete an unused var in nuget pipelines(#7345)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "aa9ab565f587f9a590dbed62cd7759c856d1b793", "messageHeadline": "FastGelu fusion for Megatron model (#7344)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 179, "deletions": 4}}, {"node": {"oid": "0da085ed480e9381a529992fd1568f37ec87a00d", "messageHeadline": "Propagate Cast operations to maximize lower precision (float16) compu\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 1176, "deletions": 7}}, {"node": {"oid": "be79575c6aaf4dc4d394d3f42455c311c8a01d5e", "messageHeadline": "Use built-in reduce_sum() for simple reduction cases, specifically re\u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 24, "deletions": 26}}, {"node": {"oid": "3eb2d349a60c08c888c9cafe6302af013243db4d", "messageHeadline": "fix typo in scenariotestscppwinrt.cpp (#7334)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "87bd836886be0e04fa83795a195b147a1d542f0e", "messageHeadline": "Fixes in symbolic shape inference (#7258)", "author": {"name": "Oliver Rausch", "email": "Oliverrausch99@gmail.com", "user": {"login": "orausch"}}, "additions": 22, "deletions": 2}}, {"node": {"oid": "75d8319286f20bebc22b7d1ebbb3949991e5a9ee", "messageHeadline": "Liqun/ort package name2 (#7337)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f62db1a09c401420e6399b7a400e8722d9883786", "messageHeadline": "quantization tools support qlinear average pool (#7309)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 135, "deletions": 0}}, {"node": {"oid": "4c862c73ed873596d2cb8b149a7bcbe94567c5f1", "messageHeadline": "for training to use new python package naming convention to explicitl\u2026", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 172, "deletions": 55}}, {"node": {"oid": "6ceee5d131d1550997c49732855318d5735484d3", "messageHeadline": "IsInf ReduceSum transform (#7188)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 569, "deletions": 171}}, {"node": {"oid": "f8a36dd6b32123c9b8d2a46e5bcd580c8179bf65", "messageHeadline": "Add DropoutGrad function body (#7310)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 254, "deletions": 54}}, {"node": {"oid": "a5d3a52d1a8f09560982497bb5482f58559e705d", "messageHeadline": "Add Tile grad (#7289)", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 140, "deletions": 19}}, {"node": {"oid": "ce9cd6ad9acea318cfb411dff3216f332a8f4026", "messageHeadline": "Update usage of generator expression $<COMPILE_LANGUAGE:L1,L2> which \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "2c96050336676aa01f3d59a920cf06094428609a", "messageHeadline": "Fix SDL warning (#7331)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "f34468a3099627e016fbb9f87d9c09827c206180", "messageHeadline": "Fix TRT EP memory leak (#7195 revisited) (#7276)", "author": {"name": "Ahmad Zakaria", "email": "zakaria.mohammad@tomra.com", "user": null}, "additions": 13, "deletions": 14}}, {"node": {"oid": "f616ea632e3d0ceb9b1cc961fb305d93e25b4e8c", "messageHeadline": "remove mlas unittest.cpp which is already refactored. (#7319)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 0, "deletions": 3430}}, {"node": {"oid": "fce67e2b9be965e0b5dc98012f1cb477aa15f002", "messageHeadline": "Create Android Package pipeline (#7295)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 119, "deletions": 26}}, {"node": {"oid": "b7c89ce78ae0be02cb989e864e38eb8e955ee91f", "messageHeadline": "User/sheilk/add api usage telemetry (#7320)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 26, "deletions": 0}}, {"node": {"oid": "8cafd2fb2c87455e6385a599e154f6b170037cd5", "messageHeadline": "Merged PR 5918130: Add activation fusions missing in newer opsets", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 57, "deletions": 0}}, {"node": {"oid": "4971310d6aaf90452a72856670db4b5e6b74e0e6", "messageHeadline": "Fix split op in the way it deals with the optional input (#7302)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 32, "deletions": 10}}, {"node": {"oid": "61ba9ac1bb06f1e65e5985417b48efc36ab0851c", "messageHeadline": "matmul in dnnl (#7311)", "author": {"name": "jeyblu", "email": "kiefer.kuah@intel.com", "user": {"login": "jeyblu"}}, "additions": 374, "deletions": 21}}, {"node": {"oid": "21c282ed542e7673d1eea0f54e113b84c40d5c50", "messageHeadline": "yolov3 accuracy (#7235)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "6334c292408a1df4d19835a1edd1d563d932880b", "messageHeadline": "Zhalei/mlas test (#7213)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 4336, "deletions": 2}}, {"node": {"oid": "75c0192e4fb05ffbcbaf9bb96ec16bca89151010", "messageHeadline": "enable more unit tests for ROCM EP (#7307)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 73, "deletions": 30}}, {"node": {"oid": "f27f5afd8a9e7099177826befb113640280e108a", "messageHeadline": "NCHWc: Support \"sizes\" argument for Resize transform (#7290)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 108, "deletions": 42}}, {"node": {"oid": "2edf29552ddd4fe5739c0f31a331f9fd45e16254", "messageHeadline": "Add Optype to type mismatch message (#7305)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b221a4fd868409abf7cbd1fb98aa8f584e03a23f", "messageHeadline": "Better error message when ORTModule used with torch.DataParallel (#7287)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 19, "deletions": 0}}, {"node": {"oid": "c22963c23d4ff5a71a23ea8a8479424117af4a89", "messageHeadline": "Polish Lamb Kernel (#7299)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 76, "deletions": 84}}, {"node": {"oid": "711cc99f4d3ad0eaec8061b1afbd6345d5585083", "messageHeadline": "Improve logged message for nodes that are forced to execute on CPU ra\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "8ad5007f8fc87438c8e9f0e62e5873408533638b", "messageHeadline": "Polish Adam kernel (#7294)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 63, "deletions": 495}}, {"node": {"oid": "274e2fea0cd84a4cc039f3026ac87563bec24b16", "messageHeadline": "change half gemm to use compute_32f as default (#7253)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 344, "deletions": 160}}, {"node": {"oid": "a4fdb4dbd9ce1336d4d35b7ae784df04e93c1cf1", "messageHeadline": "Support transpose by merge Reshape etc into direct xint8 operators. (\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 419, "deletions": 49}}, {"node": {"oid": "42051c912ad2716cbe16e93d0850b89bffa8397f", "messageHeadline": "Narrow profiling scope (#7281)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 54, "deletions": 25}}, {"node": {"oid": "370f9b88c22723873b1961c6cd15fb628a0c1655", "messageHeadline": "Enable CoreML EP for minimal extended mode (#7266)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 69, "deletions": 18}}, {"node": {"oid": "7b4362c21a7ad6d242babb98a6f081a248cd657d", "messageHeadline": "Add support to dynamic positional/keyword input for ORTModule (#7189)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 446, "deletions": 304}}, {"node": {"oid": "4969431eba64bbd13cb701ebdd0a224935bd4f9f", "messageHeadline": "Fix codeql java warning (#7280)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "0d49e53985e2494f3b9caf4db9e8646293d1cd6c", "messageHeadline": "[Symbolic shape infer] fix scalar shape in Expand (#7285)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bc6ef809bb81046e6922e6eef496fce567dc4ecd", "messageHeadline": "NCHWc: avoid buffer reordering around Add nodes (#7279)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 189, "deletions": 55}}, {"node": {"oid": "e14b291ce7603f786e90c6ba5ac41b2532999601", "messageHeadline": "Enable symbolic shape inference in ORTModule (#7282)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "d272c8434d9943d4a779e1a225534538d3c378e9", "messageHeadline": "Suppress tracer warnings from onnx export in ORTModule (#7221)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 59, "deletions": 17}}, {"node": {"oid": "27e778909d75d0d7ff1177f3e86ef8c19db71d5d", "messageHeadline": "[OpenVINO-EP] Enabling save/Load blob feature (#7054)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 267, "deletions": 69}}, {"node": {"oid": "def4cc09c7f3c4d532289a9e6ac0eef4a6d28f9e", "messageHeadline": "Add QGEMM benchmark (#7268)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 134, "deletions": 17}}, {"node": {"oid": "aa2c46514323891e9ca2a52a5f51094c5390ccdc", "messageHeadline": "Restrict ConvGrad to __CUDA_ARCH__>=700 (#7278)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 12, "deletions": 3}}, {"node": {"oid": "beb299e17dead7d86199fac6c090c779c7e27a79", "messageHeadline": "ConvGrad CUDA Kernel Bugfix (#7273)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 9, "deletions": 4}}, {"node": {"oid": "844361bc67dcd149987290ec326dc492376190fb", "messageHeadline": "Support eval mode and torch.no_grad context in ORTModule and restruct\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1019, "deletions": 562}}, {"node": {"oid": "025abf996dfce0d9c7e2957e0efcc28a936454f3", "messageHeadline": "fix for using tensorrt:20.12 base image (#7264)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "4bc17ca04eccfb15fa2db0b0f1b4f5bd6d369a47", "messageHeadline": "CUDA ConvGrad Kernel (#7227)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 386, "deletions": 30}}, {"node": {"oid": "8219518aa8d2a1dc2fe9dec4a24358200e52de4e", "messageHeadline": "Fix initializer counts when used as graph output (#7260)", "author": {"name": "TomWildenhain-Microsoft", "email": "67606533+TomWildenhain-Microsoft@users.noreply.github.com", "user": {"login": "TomWildenhain-Microsoft"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "2ec452cdad30355c0f138a53a3360a3bf5a2c2d9", "messageHeadline": "Remove ROCM workaround for half-to-double cast.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 1, "deletions": 9}}, {"node": {"oid": "25e261f1961b968acb1d6e25f18699729707ccc1", "messageHeadline": "Avoid passing zero bias to Gemm in gradients (#7244)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 12, "deletions": 26}}, {"node": {"oid": "405ca49012dc7cf7c225660174053aaa9ee52e87", "messageHeadline": "build ONNXRuntime into WebAssembly (#6478)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 3562, "deletions": 142}}, {"node": {"oid": "2aa89989c41dcf036eec18a38133b8917b9626df", "messageHeadline": "Not-where fusion (#7182)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 261, "deletions": 0}}, {"node": {"oid": "790fc11e6012c6591a258cea15c56c4c87685b54", "messageHeadline": "QDQ: type conversion and more ops support (#7243)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 784, "deletions": 108}}, {"node": {"oid": "5d759e182b30f152ad87620371af58660e2a8b77", "messageHeadline": "Allocate external Rocm allocator via PyBind (#7148)", "author": {"name": "raviskolli", "email": "48601275+raviskolli@users.noreply.github.com", "user": {"login": "raviskolli"}}, "additions": 276, "deletions": 76}}, {"node": {"oid": "6308e709cc392e2b940a75357fac88021f24e47a", "messageHeadline": "Update opset for other training graphs to 12. (#7259)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "a9ff4c29e5a4fadab9124bfb07a491820432ebd3", "messageHeadline": "Add function body to GeluGrad schema (#7190)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 62, "deletions": 1}}, {"node": {"oid": "dbcfc4bee63d61a70c31de82b796a4d700ee96d6", "messageHeadline": "Add mlas_bench tools. Starting with sconv bench and sgemm bench. (#7139)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 462, "deletions": 0}}, {"node": {"oid": "56b22c1c6b8cc8c9b541436c154e9be1a5cbbc3b", "messageHeadline": "Fix assert that the tensor's device type is 'cpu' #7248", "author": {"name": "ashari4", "email": "70242157+ashari4@users.noreply.github.com", "user": {"login": "ashari4"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e9ffcfa24741e4a37dae161b4c81e09b75187156", "messageHeadline": "Add cuda kernels for GreaterOrEqual, LessOrEqual, Where; modify Clip \u2026", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 376, "deletions": 33}}, {"node": {"oid": "c85657cfd7100c9de7b4a062d45082f8689061d1", "messageHeadline": "Update test_training_model.onnx to opset 12. (#7251)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "a9dbb511fb1fcf28ade60604f3c59f4ba4958318", "messageHeadline": "MLAS: fix qgemm bus error with Android + ARM32 (#7250)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 28, "deletions": 30}}, {"node": {"oid": "fb40602ea2b85ca276417dc88b935f6a05124371", "messageHeadline": "Mem trt (#6868)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 540, "deletions": 293}}, {"node": {"oid": "2fcd69d644f4fb9a8e4eaa2981c0e74b32577b39", "messageHeadline": "Cleanup build.py (#7245)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 29}}, {"node": {"oid": "5bd192c439a8b3e6e53bd46df0989eb4850c8040", "messageHeadline": "Update ContribOperators.md (#7246)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 410, "deletions": 13}}, {"node": {"oid": "3b16afc0db90a676efef3ce270978a406b3855fa", "messageHeadline": "Make dW optional for convgrad (#7083)", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 74, "deletions": 66}}, {"node": {"oid": "c5973fbbacde13fad9001a38fab7287ae3f801f2", "messageHeadline": "Update the build script for Android AAR package (#7229)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 24, "deletions": 16}}, {"node": {"oid": "9f14af9809c328c76764675c6dad05c83d52e950", "messageHeadline": "Add BERT-L perf regression test on MI100 and re-enable batch size tes\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 63, "deletions": 30}}, {"node": {"oid": "10102c09b67a1f1dd71e035a768fdeb86a1e85f2", "messageHeadline": "Add better model test error messaging (#7239)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 4, "deletions": 0}}, {"node": {"oid": "e7c5dcd57211f73c6da9eac7fe8b071254a4b320", "messageHeadline": "Fix Zip-Nuget-Java Packaging Pipeline (#7208)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 13, "deletions": 1}}, {"node": {"oid": "3ee9b0ec4dfa1005680c00460149b6d941803a9f", "messageHeadline": "Add detailed assertion error message (#7232)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "008065aab16e46d37b3847a4759ff0e6d5963b18", "messageHeadline": "Update README.md (#7043)", "author": {"name": "Marek \u0160uppa", "email": "mrshu@users.noreply.github.com", "user": {"login": "mrshu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2b8513539ebbb0e201538875c0e81e73563ef441", "messageHeadline": "Div mul fusion (#7183)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 226, "deletions": 0}}, {"node": {"oid": "74ee24cf7f4810b056d04ba19bc5acc48e939f1e", "messageHeadline": "rename cuda_mem_limit and hip_mem_limit to gpu_mem_limit for both CUD\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 60, "deletions": 64}}, {"node": {"oid": "68b12a61794474ae36056585b62d77dc5a921e81", "messageHeadline": "Support for saving and loading pytorch compatible state dictionaries \u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 169, "deletions": 0}}, {"node": {"oid": "8d737f977056444a307f1b7f0bcd402fba62d790", "messageHeadline": "handle optional input in quant topo sort (#7223)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 86, "deletions": 4}}, {"node": {"oid": "59b57d8322b98eb9f5d6d305aeb3c8314940fd9a", "messageHeadline": "HSA_NO_SCRATCH_RECLAIM and RCCL_ALLTOALL_KERNEL_DISABLE are not neede\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "ba5f056b0901e0e831c20ec3ab1e3412723e8a7a", "messageHeadline": "move trt_profile to TensorrtFuncState and reuse it (#7195)", "author": {"name": "Ahmad Zakaria", "email": "ahmadz1991@gmail.com", "user": {"login": "AhmadZakaria"}}, "additions": 15, "deletions": 16}}, {"node": {"oid": "ef88dc912c6a6f192e82d2c4936f094857ec34dc", "messageHeadline": "enable more unit tests for ROCM EP (#7222)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 39, "deletions": 20}}, {"node": {"oid": "afbbeaa30aebdfc9451779013fe91ea58f7e85ba", "messageHeadline": "[NNAPI/CoreML EP] Add Onnx opset 14 support (#7211)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 41, "deletions": 12}}, {"node": {"oid": "a98c2ebb8c0c77f4559e702e1ac7985af06cb2c0", "messageHeadline": "Enable saving optimized models in OrtModule (#7214)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 29, "deletions": 6}}, {"node": {"oid": "ebde3209502b753a10eb5f5d245161564b65b396", "messageHeadline": "Add cupti path for python gpu packaging pipeline (#7200)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "2d352056cf050785f672d32ee3df86e1228fb040", "messageHeadline": "Support SkipLayerNorm for ROCm EP (#7210)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 146, "deletions": 12}}, {"node": {"oid": "a3f17c8b0d74e6033f083cc01535994059a6022e", "messageHeadline": "update lamb and GatherGrad kernel for ROCm EP (#7184)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 31, "deletions": 1085}}, {"node": {"oid": "17f91ff41065c80f064ac5a90e4bd3c8c19fee5d", "messageHeadline": "remove un-needed header file. (#7193)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 0, "deletions": 4}}, {"node": {"oid": "5a6d47762531ed0b108cdce50dbe81ff3b3863e3", "messageHeadline": "Make IDataTransfer be directly shared with shared providers (#7215)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 15, "deletions": 11}}, {"node": {"oid": "0ebeaf529dca902173ac7f241067c5a128caf91b", "messageHeadline": "Check kernel def hashes (#7120)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3474, "deletions": 76}}, {"node": {"oid": "48fcddd2c5d758a7fcc70a891021f18b7ddc9d72", "messageHeadline": "Merged PR 5873494: Resize support nearest_mode floor in DML EP", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 73, "deletions": 20}}, {"node": {"oid": "15c67ddbf01f4398f28962d4334b00b773a77e7d", "messageHeadline": "Make output 1 of ConcatTraining Optional and place on CPU (#7199)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 37, "deletions": 14}}, {"node": {"oid": "45434599840ac8125c384afed92e5dd6d90c4ead", "messageHeadline": "MIOpen supports MIOPEN_REDUCE_TENSOR_AVG now.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 1, "deletions": 10}}, {"node": {"oid": "34a8b22186fae6edf433ba7861cc463e42ac0288", "messageHeadline": "disable prepacking in training (#7201)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "52bcef4d4f8573f73b17fc3981f63e3d6fcb9b52", "messageHeadline": "Openvino ep 2021.3 (#7180)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 252, "deletions": 116}}, {"node": {"oid": "249a2c14effe546c0487d5793325e7fedaf68bc0", "messageHeadline": "Pin version of pytorch to 1.8.1 for ORTModule CI pipeline (#7167)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 5, "deletions": 6}}, {"node": {"oid": "fc6ac5bfacc4a7fafc6bc5625c08b3d4e188af92", "messageHeadline": "dnnl fixes (#7202)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "329fd03bb4458e77a08288457479a284f783bd3c", "messageHeadline": "Add int32_t as required type to some operators  (#7192)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 47, "deletions": 25}}, {"node": {"oid": "9a8991e9b65ff3a5a5719f4d9c47669729ba5d86", "messageHeadline": "Merged PR 5866671: Move onnxruntime arm64x forwarder", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 31, "deletions": 8}}, {"node": {"oid": "04679e31ab40b2f27f70f20e65b84365066142c4", "messageHeadline": "Specify CUDA compute capability 7.5 in Linux GPU build (#7203)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0e0dd50e395afa0eaf99ed8988fdfca11146e3c3", "messageHeadline": "Support int32 type for TopK CPU op (#7089)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 19, "deletions": 1}}, {"node": {"oid": "057de97d92a146f5e1d676d51110dee6652cd246", "messageHeadline": "Merged PR 5866812: Decompose unsupported QLinearSigmoid operation in \u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 118, "deletions": 8}}, {"node": {"oid": "56d2c4baa205225a91edb659c61acfb74c8bd753", "messageHeadline": "Merged PR 5861108: Allow nodes in DML graph partitions with empty sha\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "b370ddbf5ee823d0c1e5e095fd03bf6a290e7b06", "messageHeadline": "Removes unnecessary transpose in operator Einsum (#7141)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 78, "deletions": 12}}, {"node": {"oid": "d500c5952b6fd39ba27a06bbb776623f30347650", "messageHeadline": "Add Android AAR packaging script for ORT-Mobile (#7138)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 326, "deletions": 10}}, {"node": {"oid": "0fdef1bf470db952c92e7f82a5a2e4e83ef3c685", "messageHeadline": "[Node.js binding] upgrade y18n to v4.0.1 (#7185)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 8871, "deletions": 9231}}, {"node": {"oid": "45cb0cae8c7c2de6e8b1dcbab3651dc0f7e0e891", "messageHeadline": "Adding TorchEmbedding contrib op (#7136)", "author": {"name": "Negin Raoof", "email": "neginmr@utexas.edu", "user": {"login": "neginraoof"}}, "additions": 301, "deletions": 14}}, {"node": {"oid": "e54560449960527251540d35d1cea32424a14e5d", "messageHeadline": ". (#7165)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d880578537fa9c3d78b15dad7b3818446f09e7e8", "messageHeadline": "Exclude cpuid.h from Mac non x86 arch (#7166)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "0ccfe6c86a144248b9ec0d6a78efa42bfe37cc16", "messageHeadline": "Enable type reduction for Scatter/ScatterElements CPU kernels (#7171)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 163, "deletions": 53}}, {"node": {"oid": "07201bac7a68d385b4817c9804398a7bc597f207", "messageHeadline": "expose session option and provider options (#7112)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 90, "deletions": 18}}, {"node": {"oid": "c4ebc608702c001b53d54ae32d05d3deb194932d", "messageHeadline": "sort quantized nodes in topo logical order (#7172)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 50, "deletions": 3}}, {"node": {"oid": "4f303412539bcc51f0d8a09e2a510b799afe558d", "messageHeadline": "Check the count of DequantizeLinear for matmul (#7174)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 34, "deletions": 1}}, {"node": {"oid": "a01334ba5657bce9ed51cebd76272ad223574507", "messageHeadline": "MLAS: activate udot kernel on Windows ARM64 (#7169)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 20, "deletions": 7}}, {"node": {"oid": "bbcf419ac61b1a619e636e0d88419f23c447495b", "messageHeadline": "Move the Windows GPU machine pool of Onnxruntime packaging pipelines \u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 16, "deletions": 255}}, {"node": {"oid": "d1acdd4f4b1c9983c2b8064d8d5bf6c5fe1e6b5d", "messageHeadline": "Support building ARM64EC onnxruntime.dll (#6999)", "author": {"name": "Ben Niu", "email": "niuben003@gmail.com", "user": {"login": "mcfi"}}, "additions": 28, "deletions": 11}}, {"node": {"oid": "77c19436c0bfb33518406f52d8854d589209205c", "messageHeadline": "add a notebook for mobilenetv2 quantization (#7164)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 351, "deletions": 0}}, {"node": {"oid": "aeca7c29408f03ad9e19837085156cba3c117263", "messageHeadline": "Cuda Profiler  (#7110)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 205, "deletions": 35}}, {"node": {"oid": "b22e60bd442cdd9a96138a5705418a76ef46e467", "messageHeadline": "pull onnx latest commit (#7102)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 412, "deletions": 58}}, {"node": {"oid": "9297527b7ad14661480ac29cd9b855e0163d79a4", "messageHeadline": "Enable NHWC transformer when generating ORT format model (#7126)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 325, "deletions": 274}}, {"node": {"oid": "90294b9c433f8d2319bfa38e0a34caa707725b15", "messageHeadline": "Fix Transpose and MatMul fusion code to check the input datatypes as \u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 51, "deletions": 1}}, {"node": {"oid": "65ce5f07b3761f441fb8e98740f39e69cd89cc58", "messageHeadline": "add Dockerfile.rocm4.1.pytorch (#7152)", "author": {"name": "Jeff Daily", "email": "jeff.daily@amd.com", "user": {"login": "jeffdaily"}}, "additions": 197, "deletions": 0}}, {"node": {"oid": "f27835c4dee32e83ac8a4c947d64825b32580079", "messageHeadline": "Disable batch size test for AMD CI pipeline after agent upgrade to Ro\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 14, "deletions": 14}}, {"node": {"oid": "f365f1d967edfaa942be86653c5b838d3d99e9ff", "messageHeadline": "Resize_impl.cu: Change _Round to roundf (#7140)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "63d9d5afd3b8172cb13915b8385d8a5246a4ac72", "messageHeadline": "Fix Pad and Gather incorrect usage of HasType helpers. (#7146)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 14}}, {"node": {"oid": "ab86634c365c74847497a02d3d020062b67a8ee9", "messageHeadline": "Address comments from ORTModule master merge (#7101)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 67, "deletions": 46}}, {"node": {"oid": "a8f0ab9c5f907ab3e759eb28f358ff15fa35bd55", "messageHeadline": "Merged PR 5846998: Fix warnings level for DML EP", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "a01f15198c03ba4af47fe268ae1f65a1b9373f77", "messageHeadline": "Add support for large models (#7113)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 73, "deletions": 9}}, {"node": {"oid": "2b31b80b1fd549a30da1024278990cad5c20dbea", "messageHeadline": "icnrease timeout (#7145)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3771e0bf10cde8f3739dda1b1cde60ca578c4583", "messageHeadline": "update bert quantization notebook (#7137)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 114, "deletions": 121}}, {"node": {"oid": "c9b29fbd06aad25ba49f7e31c52a37f19b2f0977", "messageHeadline": "Disable MatmulTransposeFusion for CPU EP (#7135)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2bf54bcaa2e299af56f8d4eeb8b94621cac14ac1", "messageHeadline": "Fix bugs in sparsify script (#7134)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "cc0e7bee76bbd4fece212182c8412ebbe28c425f", "messageHeadline": "Add function-body to SoftmaxGrad (#6988)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 427, "deletions": 7}}, {"node": {"oid": "53c123dcee3660f3a44ce9519f51904bd045e763", "messageHeadline": "Add session option configuration to enable GeluApproximation (#7131)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 41, "deletions": 32}}, {"node": {"oid": "39bd192d33a010228039cfc37904e9caabb76566", "messageHeadline": "Merged PR 5837692: Merge latest from upstream", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 28079, "deletions": 14476}}, {"node": {"oid": "8e54b76e2d845b0d0f4c71101a021d1b35ec03b4", "messageHeadline": "QDQ implementation (#7033)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1387, "deletions": 1062}}, {"node": {"oid": "865c67611cb327f88adef614406cdb90d71084a9", "messageHeadline": "Exclude profiler from minimal build (#7115)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 38, "deletions": 8}}, {"node": {"oid": "fda0470683318dad8bea56e6608276b9d3790e85", "messageHeadline": "Add New AllocKind for YieldOp Outputs, Run YieldOp with InferenceSess\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 78, "deletions": 6}}, {"node": {"oid": "1c8d874412f2991f53cc3f6b16d3f25270f13097", "messageHeadline": "Promote BiasDropout from orttraining to onnxruntime (#7116)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 523, "deletions": 452}}, {"node": {"oid": "293774fbeb6d8b1ea577a20f14096c94c9fcbba2", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into p/adtsai/merge", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 28079, "deletions": 14476}}, {"node": {"oid": "cd67f12add8bece5c50454e96ccd3c4021dca83e", "messageHeadline": "Move IOBinding and RunOptions to ctx (#7028)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 66, "deletions": 32}}, {"node": {"oid": "2e3bbad19f24e5823a1a1a24df95ddc1dcb2374d", "messageHeadline": "Move TensorRT Windows CI build to the machine pool (#7127)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1c04eec2b1d8a51a37041192f98b1faf5197d1fe", "messageHeadline": "[NNAPI EP] Fix error for QLinearAdd with an initializer as input (#7093)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 175, "deletions": 103}}, {"node": {"oid": "540eac253e60a100f7bf5790fed6630382825966", "messageHeadline": "Deepspeed pipeline parallel and fairscale sharded optimizer test samp\u2026", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 428, "deletions": 1}}, {"node": {"oid": "6987106bf5580fabb390b28ff6daae3551588395", "messageHeadline": "Add missing Python dependencies for ORT training (#7104)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 14, "deletions": 3}}, {"node": {"oid": "fffe16cb43486ef6f5550f9e5c52afb02cbceb05", "messageHeadline": "Fix a bug in quant GEMM and add an unit test (#7111)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 126, "deletions": 3}}, {"node": {"oid": "b07e168a2b358e10423a29501bf49634281e6139", "messageHeadline": "Delete an unused file: download_test_data.py (#7109)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 154}}, {"node": {"oid": "5cb893445970948fc5b80f5c7d4f6a97d17e477f", "messageHeadline": "update Dockerfile for workaround for issue in RCCL for rocm4.0 (#7108)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 1, "deletions": 0}}, {"node": {"oid": "c0994fdfbb38345b2037abc2b4c7d5490243a2f0", "messageHeadline": "Update ORTTrainer to permit Rocm and permit export of opset 13 (#7059)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 14, "deletions": 5}}, {"node": {"oid": "53392664d373bb7d9c5271b686744204a34ff02d", "messageHeadline": "Enable type reduction for Shrink, Sign, SplitToSequence CPU kernels (\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 281, "deletions": 211}}, {"node": {"oid": "c3310efdcd5296d860aa269b0a0d2918ce028cc6", "messageHeadline": "Support for models having partially non trainable parameters (#7058)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 129, "deletions": 25}}, {"node": {"oid": "a7a2a16edddc283b53d7737f897b4bbda5e86209", "messageHeadline": "Pass arguments to azure_scale_set_vm_mount_test_data from perf test c\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 4, "deletions": 34}}, {"node": {"oid": "c965878a692044e117e38c96032c148d00fc9a91", "messageHeadline": "fix a bug in global average pool and add unit test (#6913)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 103, "deletions": 3}}, {"node": {"oid": "230c137460721c160c1632b6f38d99798b053321", "messageHeadline": "cmake: support install target with generated pkg-config file (#7076)", "author": {"name": "Aaron Boxer", "email": "boxerab@gmail.com", "user": {"login": "boxerab"}}, "additions": 28, "deletions": 0}}, {"node": {"oid": "309885b08de747f6de4beaaf1fbb216554d31262", "messageHeadline": "upload ort-gpu-training python nightly package to azure feed (#6998)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 28, "deletions": 0}}, {"node": {"oid": "416ee3c4d26e9535576002b2c65a88c594f8e172", "messageHeadline": "MLAS: add 32-bit transpose support (#7092)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 304, "deletions": 75}}, {"node": {"oid": "5ec0e715420091f2cc810e474ebc44732d0e48b3", "messageHeadline": "ORTModule support non-differentiable module output (#7048)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 220, "deletions": 38}}, {"node": {"oid": "be45a59d99e4e87b47553b3b1432ce1884f8370a", "messageHeadline": "Make our CUDA code be compatible with the latest VS2019 update (#7062)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "df6a68f59cb6ac01d7401047fef3066299b5547c", "messageHeadline": "Fix fallback providers for InferenceSession (#7091)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "529da3b0033961285a38ad423f1da96950b66457", "messageHeadline": "Thread pool profiler (#6748)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 366, "deletions": 27}}, {"node": {"oid": "867804bea16d15d330f1db2605f1301d94991f1f", "messageHeadline": "Add auto doc gen for ORTModule API during CI build (#7046)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 210, "deletions": 51}}, {"node": {"oid": "3b58fc7b97c9aee073c96991609f3039a85ba90f", "messageHeadline": "Add types support for Sparse Initializer in Onnxruntime (#7004)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 542, "deletions": 70}}, {"node": {"oid": "4a3d1176d709c7919ff91f79413e953cccdee1cb", "messageHeadline": "adding ngraph_DIR to fix build (#6975)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "4cbb8e166a86f13e36b56d8146515c9c4826b875", "messageHeadline": "Update kernel def hashing (#7019)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 114, "deletions": 45}}, {"node": {"oid": "06df28748f7b9a592913b55734cb1cc05cc9f57d", "messageHeadline": "Change tabs to spaces in Windows.AI.MachineLearning.idl (#7088)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "79ba045d7426670fac634cb19fedc4e11eacffad", "messageHeadline": "Enabled rocm support for graph transformations (#7057)", "author": {"name": "raviskolli", "email": "48601275+raviskolli@users.noreply.github.com", "user": {"login": "raviskolli"}}, "additions": 18, "deletions": 18}}, {"node": {"oid": "b2c6617b0f54c9410bd232acdcb0eb6e1678eacb", "messageHeadline": "Use 'as_scalar' when checking the 'cond' value of 'If' (#7063)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cec919bae991a5ba826e1f7282ce51fa30cbfa66", "messageHeadline": "handle 8 bit uint dlpack tensor (#7069)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 72, "deletions": 30}}, {"node": {"oid": "8d5bfdeb47c061f5f44e3b0642668ff5e6855fe7", "messageHeadline": "Increase timeout for Android CI pipeline by 30 minutes. (#7065)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8c3b59a0263eca2c45939fbaa46a47e757f94d9a", "messageHeadline": "Quantization calibration refactor (#6893)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 365, "deletions": 163}}, {"node": {"oid": "701e73b5b80fb3fac62d9ce216b0ac4abc063be5", "messageHeadline": "Move Linux minimal build CI pipeline to the new Linux machine pool (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "8bc275e93fa31c9698110a1f1d9665a237481be7", "messageHeadline": "Enhance Transpose, Cast and MatMul fusion when Cast and/or Fusion fee\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 115, "deletions": 56}}, {"node": {"oid": "1a1dd4843d97423ef1d322d9e2eb5ec25322de3e", "messageHeadline": "Enable opset 13 for Rocm (#7047)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 224, "deletions": 137}}, {"node": {"oid": "7c7d6debe67f4d8c8be5811bc341186f52773b13", "messageHeadline": "[CoreML EP] Add Resize Support (#7015)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 328, "deletions": 27}}, {"node": {"oid": "3d37a3c1d3fb786601be53761b34287941579c36", "messageHeadline": "Merged PR 5807585: Remove support for strided 64-bit emulation in DML\u2026", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 3, "deletions": 21}}, {"node": {"oid": "897a0b9839daee8bf64e927abbe53a6304415396", "messageHeadline": "Merged PR 5807395: Add DML kernels for QLinearAdd (com.microsoft name\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 432, "deletions": 234}}, {"node": {"oid": "514444d8207ea3accccea9fe5d90c6f84bc57cd6", "messageHeadline": "Fix pipeline generating python documentation (#7027)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 22, "deletions": 9}}, {"node": {"oid": "c60ef62190c911f63ca1e075e9d9af5c2e9cc78e", "messageHeadline": "Update ORTModule feature with remaining PRs from feature branch (#7040)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 395, "deletions": 95}}, {"node": {"oid": "4fd9fef9ee04c0844d679e81264779402cfa445c", "messageHeadline": "Support HuggingFace Models Converted From tf2onnx in Python Script  (\u2026", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 449, "deletions": 127}}, {"node": {"oid": "934bb52cfb08c35682feb0c553f1591911c14449", "messageHeadline": "Merged PR 5805461: Add ARM64X forwarder libs", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 29, "deletions": 0}}, {"node": {"oid": "bc3aea4be00bc07689bb370da69c2a3dc0b414c4", "messageHeadline": "Capitalize DLL name", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "4f5d6a0e4dbf5565f83bfebfcc028b5aedb4290b", "messageHeadline": "Add ARM64X forwarder libs", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 28, "deletions": 0}}, {"node": {"oid": "335edaa2c485ba0dec877bf4cdbd652e2d5d105c", "messageHeadline": "Merge pull request #6973 from microsoft/thiagofc/merge-ortmodule-into\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 6078, "deletions": 155}}, {"node": {"oid": "03885af5a0b7d10468552a1b200605bd9b3fda63", "messageHeadline": "Adding prepacking to QLinearMatMul (#6980)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 252, "deletions": 53}}, {"node": {"oid": "90642e7eacdf018cf26b633ec0c7ad1a1072e41a", "messageHeadline": "MLAS: more code cleanup (#7036)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 95, "deletions": 133}}, {"node": {"oid": "8e0970a0203a36a640b97e13c6b823b70a079728", "messageHeadline": "dnnl format tag fix (#6943)", "author": {"name": "jeyblu", "email": "kiefer.kuah@intel.com", "user": {"login": "jeyblu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0f9383e583d8d3429914d397886a5b756532ba76", "messageHeadline": "[NNAPI EP] Add support of QlinearAveragePool (#6915)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 206, "deletions": 77}}, {"node": {"oid": "a0fdabd23f22231c9547f1d15e2e3a061dd45dd5", "messageHeadline": "Rename all of the ONNX_NAMESPACE types for shared providers to be bac\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 510, "deletions": 513}}, {"node": {"oid": "73d085ccdd6f173c1fd031e1018aac8fadd05f4c", "messageHeadline": "add slow test (#7035)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 45, "deletions": 2}}, {"node": {"oid": "3348b8485f35939e72f7a15d614556851be69e60", "messageHeadline": "Post merge update for ORTModule", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 284, "deletions": 204}}, {"node": {"oid": "ed2d441a2ed8eff4d58353526352a70f33391213", "messageHeadline": "Update ORT server build pipeline (#7030)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 11}}, {"node": {"oid": "2e38bf5e23a9ec1db2f808532a239a3c5cbe65ae", "messageHeadline": "add TensorRT configuration to OrtProviderOptions (#6979)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 221, "deletions": 23}}, {"node": {"oid": "783acb144f04d5e70d475f6ab2445181ba62f286", "messageHeadline": "Ignored return value SDL bug fix (#6451)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "2361cb99b60e642cd22fc437842fddc53bb02b1e", "messageHeadline": "Remove CentOS CI pipeline (#6997)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 26}}, {"node": {"oid": "975e4efb8a757895d8d12ba94f220752b4da440b", "messageHeadline": "Package ARM artifacts (#6805)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3f0e50f14d056897bfdb301c4411529d6a1e1114", "messageHeadline": "Cleanup in RoiAlign (#7012)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 11, "deletions": 15}}, {"node": {"oid": "087d96200de26450ea5af6919d1185f429dd8489", "messageHeadline": "HIP_CLANG_FLAGS replaces HIP_HCC_FLAGS for ROCm later than 4.0 (#6955)", "author": {"name": "Jinsong Ji", "email": "jjsbear@hotmail.com", "user": {"login": "jjsbear"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "5480f8dd1de401c95cdd57eae89de893cf118611", "messageHeadline": "MLAS: misc cleanup (#7013)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 246, "deletions": 376}}, {"node": {"oid": "4e670f7ab1b95a27b2381cb1229e0409a1d82594", "messageHeadline": "Support larger hidden size in Attention Cuda kernel (#7002)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 273, "deletions": 63}}, {"node": {"oid": "27ac88201aafc67bbd8ee439bb9bef745a4cc549", "messageHeadline": "Support a CPU kernel for Celu (#6995)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 35, "deletions": 3}}, {"node": {"oid": "d0cca35308b484407fd326b5fe6d4a4d9bf74b78", "messageHeadline": "Add README for docs (#6626)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "e5e922ec1e9836ac928e919dc3ad57b3595ac955", "messageHeadline": "Fix some warning option override warnings from dependencies. (#6983)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 12, "deletions": 7}}, {"node": {"oid": "4c9ccb0f1af77161479237f1742ba57bb0166a59", "messageHeadline": "[OpenVino] getcapability design (#6863)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 1420, "deletions": 2770}}, {"node": {"oid": "4161758058bc98f515d349d810a31a4e128b4cc1", "messageHeadline": "Remove openmp related packaging pipeline (#6991)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 58, "deletions": 423}}, {"node": {"oid": "c588d5d13afcc3a4e1846d5c41ef59c214e6fa09", "messageHeadline": "Add rocm execution provider to prover_list (#6306)", "author": {"name": "Shucai Xiao", "email": "shucai@gmail.com", "user": {"login": "scxiao"}}, "additions": 311, "deletions": 4}}, {"node": {"oid": "031587814bc42da9a722da08d2d1ea822fc32ed7", "messageHeadline": "Add support to save onnx graph with external initializers file. (#6911)", "author": {"name": "Alberto Magni", "email": "49027342+alberto-magni@users.noreply.github.com", "user": {"login": "alberto-magni"}}, "additions": 256, "deletions": 0}}, {"node": {"oid": "12b5ab3bab7d51f61827ac7a92f7da0a4db98bf3", "messageHeadline": "Update CUDA custom op unit tests to account for recent ORT change (#6\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 472, "deletions": 393}}, {"node": {"oid": "694389a85d35668859c33915849a0c0dd3667969", "messageHeadline": "Automate generation of python documentation (#6909)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 34, "deletions": 4}}, {"node": {"oid": "f7df2f805bd080744ed33f50597da8480f42f55b", "messageHeadline": "Resolve HTTP Error 503: Service Unavailable for MNIST dataset", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "aa60a8368f52b52d0d190847682aa2eb503ddeaa", "messageHeadline": "Update type reduction operator type usage processors set. (#6976)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 8, "deletions": 6}}, {"node": {"oid": "b57a85d8638f8517ec11a1c3f237eb42413c4360", "messageHeadline": "Support symbolic shape infer in transformers tool (#6899)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 498, "deletions": 203}}, {"node": {"oid": "f4796e1953741c45522048074904f7558bd5a902", "messageHeadline": "Enable type reduction for Range, ReverseSequence, ScatterND, Split, a\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 141, "deletions": 50}}, {"node": {"oid": "4a4488baaead3479fec191fdd2d2686b6f243408", "messageHeadline": "Release buffers for prepacked tensors (#6820)", "author": {"name": "Chen Fu", "email": "chenfucs@gmail.com", "user": {"login": "chenfucn"}}, "additions": 274, "deletions": 274}}, {"node": {"oid": "50973de1a2b09b55389dfcd45c59ffe39876dc3d", "messageHeadline": "Merged PR 5691446: QLinear Graph Support", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 16, "deletions": 29}}, {"node": {"oid": "2f307dd223271b26af7508e3f125759443cb9e48", "messageHeadline": "Fix possible fd leak in NNAPI (#6966)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "89d450697b4302e4a93f1de9dc0967ebf257c514", "messageHeadline": "Introduce ORTModule training API to ONNX Runtime", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 6010, "deletions": 167}}, {"node": {"oid": "9f84819f3265142c144d3cda1ab469b2fd998750", "messageHeadline": "Update onnxruntime_perf_test.exe to accept free dimension overrides (\u2026", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 69, "deletions": 2}}, {"node": {"oid": "f723ff2285bdfbcad23f804aba470b8b6d4d1882", "messageHeadline": "fixed type to experimental session constructor (#6950)", "author": {"name": "David Medine", "email": "david.eric.medine@gmail.com", "user": {"login": "dmedine"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "4884eee642295adc88f77d9c85edd50ce4b38ed8", "messageHeadline": "Attention fusion detect num_heads and hidden_size automatically (#6920)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 279, "deletions": 30}}, {"node": {"oid": "ce403eea98fc408f66c8242e46c128e62569899e", "messageHeadline": "Add *args support for ORTModule inputs (#6883)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 106, "deletions": 40}}, {"node": {"oid": "acfe7ac4cebf313662f396faacd93e2115bad8fb", "messageHeadline": "Implement QLinearAveragePool with unit tests. (#6896)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 605, "deletions": 1}}, {"node": {"oid": "1e13e2666e61afcdc82ff17253555f5a8da65f86", "messageHeadline": "Support ROCM EP for ORTModule (#6967)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "a8b897f710a3fd8dbb2dd0229e62b7e791479f83", "messageHeadline": "MLAS: quantized GEMM update (#6916)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 6694, "deletions": 5771}}, {"node": {"oid": "bc319bd7aacb3324fc04df1484fb95e465e2009c", "messageHeadline": "Fix warning from setting multiple MSVC warning level options. (#6917)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 23, "deletions": 10}}, {"node": {"oid": "8468099f9327c46d7ad4bb9394148ad278e22ae5", "messageHeadline": "Use DLPack for Graph Inputs and External Outputs of YieldOp (#6968)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 140, "deletions": 63}}, {"node": {"oid": "3f579facbcb48eb7810116a1886b82bbd8ceee66", "messageHeadline": "Relax atol for some ORTModule UTs (#6969)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "d5ed3e7fbae11005c38876e852236bb3af753966", "messageHeadline": "Enable type reduction in EyeLike, Mod, random.cc CPU kernels. (#6960)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 238, "deletions": 149}}, {"node": {"oid": "89916fdb05a7b6cee83f49b3f72e09821eea5fc2", "messageHeadline": "fix stream sync issue (#6954)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "bdaea1d9ae24feac4907e1bdf96374dc92c9dc82", "messageHeadline": "Update baseline due to loss scale fix (#6948)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "743a93faf3dde7d7eef9180c360de05a94174154", "messageHeadline": "Fix broken link in server usage and remove absolute path from dockerf\u2026", "author": {"name": "Raduan Al-Shedivat", "email": "36044389+Raduan77@users.noreply.github.com", "user": {"login": "0xRaduan"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "534adbb06544eb8cb50f1a328b6c52863480254d", "messageHeadline": "Support ORTModule on ROCm EP (#6945)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "3b2847b2d8e1a9385775e48ad5b66ce413901ce2", "messageHeadline": "Add UT correctness and address comments for previous symbolic shape P\u2026", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 50, "deletions": 37}}, {"node": {"oid": "ba51774a1f73f77948505816d8a57d405d40bd80", "messageHeadline": "Add GPU support for DNNL endpoint (#6741)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 315, "deletions": 195}}, {"node": {"oid": "5303b33f696bd4dd641d77eb68fba02916d2bd8c", "messageHeadline": "Clean ORTModule dev branch (#6944)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 3, "deletions": 248}}, {"node": {"oid": "48eebed869d4bf9712549f84e8396ca73223da4c", "messageHeadline": "Interchange Cast and Transpose operations to facilitate Transpose-Mat\u2026", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 252, "deletions": 0}}, {"node": {"oid": "91c6a330c0d9ce7a5ff5e8bb691456d7f2d9c831", "messageHeadline": "Add UseCount for External Outputs (#6894)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 48, "deletions": 10}}, {"node": {"oid": "c8e2e3191b2d506d1260069eb3d3fc7c262ec172", "messageHeadline": "Support parsing an array of values stored as an attribute in a custom\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 272, "deletions": 8}}, {"node": {"oid": "e64eff1f131eb1ddaf8bb138752df6b2e7f205fe", "messageHeadline": "Enable build with bitcode for iOS (#6905)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 29, "deletions": 10}}, {"node": {"oid": "73fe1f2deb16f76f4b25d0813f986ab8538e2fd8", "messageHeadline": "Rename op kernel type control 'supported types' to 'default types'. (\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 187, "deletions": 182}}, {"node": {"oid": "f1ade14e442d669327e106653021495e010e9353", "messageHeadline": "Assert that the data is on the same device as ORTModule (#6942)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 37, "deletions": 0}}, {"node": {"oid": "67c67408c4328adcd5b6e14938516fa4e19c82a6", "messageHeadline": "Only set _native folder for Microsoft.AI.MachineLearning package (#6939)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "bc276521884c9ac6fb411fdffba834f8ee74cb25", "messageHeadline": "MLAS: workaround LLVM x86 assembler (#6922)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 74, "deletions": 36}}, {"node": {"oid": "b89f52c2773e27e1f6999f2b2c6aaf33f3f96380", "messageHeadline": "Add tests of Attention and QAttention for pruned model (#6914)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 127, "deletions": 25}}, {"node": {"oid": "f2f60eed59289e173a35fb369360efaae58f3b4d", "messageHeadline": "Fix broken Java API link (#6826)", "author": {"name": "Denny Abraham Cheriyan", "email": "dennyac@gmail.com", "user": {"login": "dennyac"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "15d81fb63afa0463465969b5f4138f5027cb9d9f", "messageHeadline": "Enable type reduction for Clip, MaxPool, and Pad CPU kernels. (#6918)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 172, "deletions": 61}}, {"node": {"oid": "b6c4a7ac548f56a0d35f6a2480705ca964a5286b", "messageHeadline": "Support required types when excluding typed registrations (#6871)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 273, "deletions": 178}}, {"node": {"oid": "de6e66f3d49eebec36884f488789bb429fee0c92", "messageHeadline": "Fix loss scaling when running ORTTrainer with BERT under mixed-precis\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "56c5620fd2fc1c2a15f1407d83c3a36e53b5fa76", "messageHeadline": "Disable Materializing Grads (#6822)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 206, "deletions": 141}}, {"node": {"oid": "dfc7c18e3104df7e9ed8eb4f8e012b13bbcbafa8", "messageHeadline": "Introducing TrainingAgent interface to performance training using Yie\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 285, "deletions": 192}}, {"node": {"oid": "601e04fb278c3df5d66adb3b4586f86f89e3e755", "messageHeadline": "update Readme (#6903)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "79f832c682b718b53085b90b86ef80b6df7a7678", "messageHeadline": "Separate requirements.txt file for ORTModule pipelines (#6879)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 42, "deletions": 31}}, {"node": {"oid": "ac4d6155530ebee55bcad95cfdb08a8cbe7b9b35", "messageHeadline": "Enable priority-based execution order as default to support inputs wi\u2026", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 91, "deletions": 18}}, {"node": {"oid": "9126faa35bdbc5be3add6e7a9da711174f0fc7eb", "messageHeadline": "Ability to fuse non-square (pruned) attention weights for BERT-like m\u2026", "author": {"name": "Funtowicz Morgan", "email": "mfuntowicz@users.noreply.github.com", "user": {"login": "mfuntowicz"}}, "additions": 14, "deletions": 11}}, {"node": {"oid": "f986ffcb5fb17da7e72443f3552269054a9d9988", "messageHeadline": "move pipeline file and change relative path (#6882)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 239, "deletions": 0}}, {"node": {"oid": "107c9672fd3bbf46d184b4c0626f7085ad558a6e", "messageHeadline": "No such file or directory with  --use_external_data_form and int8 (#6\u2026", "author": {"name": "Reuben Zotz-Wilson", "email": "rdzotz@gmail.com", "user": {"login": "rdzotz"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "679718b12f9f1afaae43f2f7f49d370abf1025b7", "messageHeadline": "Configure session thread pool spinning preference (#6895)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "8f1786d5d2ac5377b2a887398d9048fd06beeabc", "messageHeadline": "Save output tensors in bert_test_data tool (#6872)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 151, "deletions": 42}}, {"node": {"oid": "fa8d1b44b832ffafddee0c38ac1fe1d09d8344ee", "messageHeadline": "Fix app packaging in UWP (#6804)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 60, "deletions": 51}}, {"node": {"oid": "7915b6709a1dff27d31cdafc9528f3ba451f0e76", "messageHeadline": "Revert Gather Grad optimization in PR 6381 targeted for Rocm (#6880)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 14, "deletions": 12}}, {"node": {"oid": "54cdb6af71e50c002d32b5ac8324c47c185dfc5c", "messageHeadline": "Add check that the first 2 Loop subgraph inputs have an shape (could \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "b429edcd4591838cb88b1d14debec0f16ae2d284", "messageHeadline": "Merge pull request #6890 from microsoft/bmeswani/merge_master_onto_or\u2026", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 1953, "deletions": 1093}}, {"node": {"oid": "aa93f2e23680924f0b83cce1ebec4fce6e579baa", "messageHeadline": "move SetOutputMLValue from op_kernel.h to op_kernel_context.h", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d5667554e69b28c6db4cdfeac427a779b0ab8724", "messageHeadline": "Merge branch 'master' of github.com:microsoft/onnxruntime into bmeswa\u2026", "author": {"name": "Baiju Meswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1952, "deletions": 1093}}, {"node": {"oid": "d01006fc222799f879c8ce70edd58e97f53b7767", "messageHeadline": "Move constants from heap to stack to avoid randomness on cudnn functi\u2026", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 20, "deletions": 18}}, {"node": {"oid": "749e6a08a6a295e2b6f356e753bb881957628960", "messageHeadline": "Add more asserts for ORTModule forward's correctness  (#6887)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 38, "deletions": 21}}, {"node": {"oid": "ed1883a97c1d73810a447b09d211bcb2e34822de", "messageHeadline": "Workaround for HTTP Error 403: Forbidden for MNIST dataset (#6885)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "fedb68429cb219d8651e1c20f53d7f314154b033", "messageHeadline": "[NNAPI EP] Add per-tensor u8s8 support for Qlinear[Conv/MatMul] (#6818)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 265, "deletions": 119}}, {"node": {"oid": "3c5d811e77289632097056d5f9c9588d03952c0a", "messageHeadline": "[CoreML EP] Add [Average/Max]Pool support (#6870)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 105, "deletions": 19}}, {"node": {"oid": "9a9e741a8c0d7eefaab6b75708cd3189d2b152be", "messageHeadline": "Support optional inputs/outputs in custom op development (#6727)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 224, "deletions": 13}}, {"node": {"oid": "f22f04a109b4e0ffa592f50a49f0f2b4569e4cfa", "messageHeadline": "Add comment (#6860)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 19, "deletions": 38}}, {"node": {"oid": "6285ee2398f2fe448e84e69158b4cd2300c0c6d2", "messageHeadline": "Reroute quantization tool readme to /docs page (#6854)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 2, "deletions": 147}}, {"node": {"oid": "9073f7a5c3e7f5147682567f7217cd1a63c8fd2e", "messageHeadline": "support opset13 in embednorm (#6866)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 99, "deletions": 28}}, {"node": {"oid": "0d0eb2c85ca399d1d5a39f4f11edbb34fdb78756", "messageHeadline": "Change OpKernel class to be shared with shared providers (#6837)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 450, "deletions": 449}}, {"node": {"oid": "38796ad45125b95ab72d95ab737b3480e534a367", "messageHeadline": "Refine force CPU fallback logic in the CUDA EP  (#6849)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "4238ce341a2389a2ced50430afbcf748610b22a5", "messageHeadline": "Add External Outputs Flag for YieldOp (#6789)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 115, "deletions": 2}}, {"node": {"oid": "66df167a73ca14431612611677a71defc8e9288b", "messageHeadline": "Add support for op kernel type control required types, require int64 \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 335, "deletions": 83}}, {"node": {"oid": "36a44d55edc8b681c174774bf9f5276350ad5070", "messageHeadline": "Only report Android Baseline binary size for master branch (#6844)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "5cf66069645d338b481e70dcfe869161c03e102f", "messageHeadline": "[CoreML EP] Add Concat support (#6834)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 189, "deletions": 0}}, {"node": {"oid": "12edf22f116578473d2cf599b2367564757948e7", "messageHeadline": "Merge pull request #6838 from microsoft/mzs/ortmodule-api-sync-from-m\u2026", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 13907, "deletions": 12422}}, {"node": {"oid": "2d6e10ba0054bfbbbc89f1e6d945b4ca3d59631f", "messageHeadline": "Update Attention and QAttention to support pruned model (#6819)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 247, "deletions": 188}}, {"node": {"oid": "f71d93ea2b7040a9d23e45b945af511a473f1daa", "messageHeadline": "Enable PyTorch Lightning basic test on CI (#6809)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "ca48310d6d72031ca9d5c3126271f43c3b1f650b", "messageHeadline": "Merge branch 'master' of https://github.com/microsoft/onnxruntime int\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 13907, "deletions": 12422}}, {"node": {"oid": "cb8d8464bc6ee39a894ecc76e05574095f5eb489", "messageHeadline": "Do not create compute stream when external CUDA allocator is used. (#\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 26, "deletions": 10}}, {"node": {"oid": "b4b87ac7a00585fff27632ac2faf653cdbc95b9f", "messageHeadline": "update (#6827)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 11, "deletions": 14}}, {"node": {"oid": "059ed1c2418be171f6c1c053ffd941273abbf2d3", "messageHeadline": "Copy forward signature from PyTorch model. (#6777)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 134, "deletions": 119}}, {"node": {"oid": "c1b0cf6d0b05a7276d7a497d2347628f986e2f9a", "messageHeadline": "Add pipeline to clear the cache for huggingface transormers models (#\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 25, "deletions": 0}}, {"node": {"oid": "355057cf9ce34eb57c9be1f644b403aa94830eb8", "messageHeadline": "Added RequiredGrad attribute to YieldOp (#6657)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 142, "deletions": 38}}, {"node": {"oid": "d5175795d2b7f2db18b0390f394a49238f814668", "messageHeadline": "Improvements to quantizer: Removed unused qType field, add reshape op\u2026", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 145, "deletions": 28}}, {"node": {"oid": "34261087396f31b74c4874665928efa474651832", "messageHeadline": "Fixed issue in python cmake to update wheel package (#6384)", "author": {"name": "Surya Siddharth Pemmaraju", "email": "surya.siddharth.pemmaraju@intel.com", "user": {"login": "suryasidd"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "8a450d523f3e4a4b8abfb3fd1f7f7053e72bc1b4", "messageHeadline": "Check gradient correctness in the UTs (#6803)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 152, "deletions": 98}}, {"node": {"oid": "fa8a9015bd74a166f1517cd61f933353349a4f8c", "messageHeadline": "Mount hf model cache and use cache for loading hf models (#6810)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 24, "deletions": 8}}, {"node": {"oid": "99ffffbe6ad3dbf639ee519d8b12ec845c7a9bf6", "messageHeadline": "Remove backward workaround from test. (#6811)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 1, "deletions": 4}}, {"node": {"oid": "46e026e9004b499daf10e95f478b828018d35d3e", "messageHeadline": "Merged PR 5727374: Github to DmlDev Update  2/25/2021", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 3669, "deletions": 2985}}, {"node": {"oid": "a134b1f8081c45b1af8dcb8084fce1cb594f18dd", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into dmldev_temp", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 3669, "deletions": 2985}}, {"node": {"oid": "b05403d8773b4c76dfa0080e6c00652243a43127", "messageHeadline": "Clear iobinding outputs (#6774)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "9b3171e95cf767bcf11bd7507cadb092abe295ce", "messageHeadline": "Make keepdims to its default value when adding ReduceMin/ReduceMax fo\u2026", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 14, "deletions": 10}}, {"node": {"oid": "db05d53b94d8c67817da5c0b08aba935c941a444", "messageHeadline": "Setup perf in docker and add features (#6582)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 379, "deletions": 207}}, {"node": {"oid": "d5f292ab7331377c525ec9fdb075b6095296b507", "messageHeadline": "fix issues caused by quantize/calibrate changes (#6802)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 10, "deletions": 7}}, {"node": {"oid": "7465673e3306f5f2e5ed670c5a18468824daf6fa", "messageHeadline": "[OpenVINO-EP] Find package changes (#6801)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 7, "deletions": 55}}, {"node": {"oid": "8a148e44fb485bb84be935f0904cc1876b38a8ee", "messageHeadline": "make ci pipeline also run batch and convergence test (#6798)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 24, "deletions": 0}}, {"node": {"oid": "ab1713f5ccdd643603b30fb7d526b2cf69db9c45", "messageHeadline": "Fix regression in constant folding optimizer (#6795)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 32, "deletions": 2}}, {"node": {"oid": "8e200e13fe76270e17bdfb6bde7b699ca46ce8fb", "messageHeadline": "Rewrite ORTModule background task coordination  (#6700)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 471, "deletions": 171}}, {"node": {"oid": "7ce4075bbda51fb00180559cad4e563b3ad1d1f9", "messageHeadline": "Support nested sequence and mapping types in ORTModule (#6791)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 367, "deletions": 145}}, {"node": {"oid": "40fa40f3ce7419a093467c70dc8ff8fcfed9335f", "messageHeadline": "Enable more unit tests for ROCM EP (#6776)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 317, "deletions": 119}}, {"node": {"oid": "aa5cd37ac84638660d7b5c99f47f32a05dca9d41", "messageHeadline": "Refactor device handling and basic support for PyTorch Lightning (#6758)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 240, "deletions": 106}}, {"node": {"oid": "f4acdb2ecd254b17b3156baaba723873e2ef8c8a", "messageHeadline": "Update transformers benchmark for transformers 4.3.* and ORT 1.7 (#6796)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 137, "deletions": 316}}, {"node": {"oid": "71a70ecf6e57ae5232d9ead899f9d1cfacd9554b", "messageHeadline": "Allow 3D ConvTranspose in CUDA execution provider (#6794)", "author": {"name": "fthielke", "email": "fthielke@fb3.uni-bremen.de", "user": {"login": "fthielke"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "5a473216b7e0143dd2b59a471aec8384afcf5c05", "messageHeadline": "[Java] Adds extra providers (#6770)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 214, "deletions": 4}}, {"node": {"oid": "47c8e9ad28e31c1e6f4e693e42d31857612e10a7", "messageHeadline": "Adding fp16 support for Einsum Cuda kernel (#6775)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 232, "deletions": 4}}, {"node": {"oid": "c02ec38f8aee534b29a09d9508d4115c6ee91189", "messageHeadline": "[Running CI now] Remove duplicate tests to speed up CI (#6768)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 0, "deletions": 80}}, {"node": {"oid": "7aa69a1f2a3de3c576ffbe171cab7014e3a741ee", "messageHeadline": "Support building python bindings when building unit tests is manually\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 85, "deletions": 77}}, {"node": {"oid": "65ba51d93eed1e17d4b7ff48c98f571e5a6c34f2", "messageHeadline": "Re-enable test and increase timeout (#6785)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "563218dcda709d0dfebadd5f2cdce37e75474929", "messageHeadline": "Update torchtext usage for pytorch transformer sample (#6767)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 91, "deletions": 64}}, {"node": {"oid": "b8b41e3775aede3b6883c5f4297149615c7acefa", "messageHeadline": "Update DirectML 1.4.1 to 1.4.2 for ORT 1.7 (#6780)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "09a5d6a9dcfc3030ee7f526e826d968823373af8", "messageHeadline": "Update docs/ONNX_Runtime_for_Mobile_Platforms.md with info about op t\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 30, "deletions": 14}}, {"node": {"oid": "5db0c9c64886d2a4372221f848f86be536793074", "messageHeadline": "Enable CI to cover globally allowed types (#6778)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 42, "deletions": 3}}, {"node": {"oid": "8703e2c778e3fbd3abf7b25e07d57fc5b5a692a1", "messageHeadline": "update benchmark_longformer for default test suite (#6772)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 197, "deletions": 155}}, {"node": {"oid": "8a6f6bc38b17febb15d375fc7de22a4f3e798843", "messageHeadline": "add --enable_cuda_line_info to build.py (#6773)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "f5313cc4ce0e4785aae2339433125748d7c69528", "messageHeadline": "[CoreML EP] Add options to enable CoreML EP only on hardware with App\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 59, "deletions": 1}}, {"node": {"oid": "ee35be012920f7afb6a7eed9728091eca32e0105", "messageHeadline": "Support specifying globally allowed types from build script (#6677)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 332, "deletions": 204}}, {"node": {"oid": "c91f3142171f6d3dad944d2e107ee3d02e8f39db", "messageHeadline": "Add robust dependency check for Python package (#6436)", "author": {"name": "Ivan Stojiljkovic", "email": "17503404+ivanst0@users.noreply.github.com", "user": {"login": "ivanst0"}}, "additions": 85, "deletions": 30}}, {"node": {"oid": "9f7dffb109a936981ca679accedbf61ccf785379", "messageHeadline": "Make it easier to fold MakeString usages to reduce binary size. (#6754)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 47, "deletions": 3}}, {"node": {"oid": "3bda7f4d36ba11c81b02e82faf363c9ce04f650d", "messageHeadline": "Fix longformer parity and perf regression (#6760)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1449, "deletions": 390}}, {"node": {"oid": "58f3aca95d798b18cf27617d9009db958e2e5ee2", "messageHeadline": "Support keyword arguments for ORTModule (#6539)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 101, "deletions": 47}}, {"node": {"oid": "47519623cd594b0ed6bf9b9816fb199ce8a757fd", "messageHeadline": "Fix iOS/macOS build warning for inconsistent symbol visibility settin\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 25, "deletions": 13}}, {"node": {"oid": "c170061998e1adfb5a37dfe87d2f845b202cd348", "messageHeadline": "Removed BUILD.md from master as source now lives in gh-pages (#6709)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 8, "deletions": 1291}}, {"node": {"oid": "20d6613efb00e5648c734b46c0a36ca3e30580b4", "messageHeadline": "Add direct link to build instructions on readme (#6729)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "a25c8e52a9b502fca6bc9838e016d55192185277", "messageHeadline": "fix link to samples in nodejs README (#6746)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "67e7a9c02db70574db1777e2b57e8d34fdb0585f", "messageHeadline": "Merged PR 5703823: GitHub RI OnnxRuntime to DmlDev", "author": {"name": "Brian Martin", "email": "brianma@microsoft.com", "user": {"login": "martinb35"}}, "additions": 11044, "deletions": 10202}}, {"node": {"oid": "79b966b01ac1960484f5330fe8fe92d60c881ed5", "messageHeadline": ". (#6751)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 4, "deletions": 62}}, {"node": {"oid": "609d0ec953159dd70e1ed2c25198b8c71153ae4b", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into dmldev_temp", "author": {"name": "Brian Martin", "email": "brianma@microsoft.com", "user": {"login": "martinb35"}}, "additions": 11044, "deletions": 10202}}, {"node": {"oid": "1a2f1bd23a09e2c9239c8a2e3c8384a33ee4a5dc", "messageHeadline": "Enable external CUDA allocator in ORTModule. (#6745)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 30, "deletions": 11}}, {"node": {"oid": "39d182f7fc25ec7ded80e8c4397ae675832037ca", "messageHeadline": "ORTModule - FastGeluFusion/fp16 fix and minor LayerNormFusion cleanup\u2026", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 194, "deletions": 5}}, {"node": {"oid": "fb3f1f5cc133d3b86d5cc412abb2e8a6fe963207", "messageHeadline": "Enable custom ops on ORTModule (#6740)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 42, "deletions": 4}}, {"node": {"oid": "67c478ede453fde70fcbd0b3f880aab419afd682", "messageHeadline": "Entropy method for calibration-based quantization (#6619)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 705, "deletions": 91}}, {"node": {"oid": "3722dd26925ac0915c731be0cdd6a722c0cd47ca", "messageHeadline": "Update docs MCR image publishing policy change (#6743)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "53eb948f4c4524565ade8af4d0171bef226363d1", "messageHeadline": "Upgrade TensorRT to v7.2.2 (#6452)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 79, "deletions": 29}}, {"node": {"oid": "b7b56121594b801d066ffed622dc57fd7072ede5", "messageHeadline": "Merge pull request #6742 from microsoft/mzs/sync-from-master", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 3217, "deletions": 1926}}, {"node": {"oid": "b1a12b49b71a24d3236895ab0fa74638ab63d0da", "messageHeadline": "Avoid removing constant weight that is graph output (#6735)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "ea3aee4d5f946f82b64ef64092651ab702ca38e9", "messageHeadline": "Bumping up version to 1.7 (#6736)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 9265, "deletions": 9260}}, {"node": {"oid": "40dda452cf7067a91ece81d3d169fa5dae3cf353", "messageHeadline": "Merge branch 'master' of https://github.com/microsoft/onnxruntime int\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 3217, "deletions": 1926}}, {"node": {"oid": "e44ac6524f8d75da5c68c99aeebc5c8786f2bfb3", "messageHeadline": "Plug n Allocate with external CUDA allocator via PyBind. (#6679)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 858, "deletions": 774}}, {"node": {"oid": "dd8ef4409ae6cbb7bf92fe9858bf56844ccbe787", "messageHeadline": "Liqun/migrate perf test (#6733)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 45, "deletions": 3}}, {"node": {"oid": "2c5e603bad81c2028a04cf0fcd250210deb13396", "messageHeadline": "Liqun/nuphar nuget (#6656)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "21f9e32c60aacffd13dcbfc0ffe1d129026be447", "messageHeadline": "Merge pull request #6714 from microsoft/thiagofc/merge-from-master", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 23387, "deletions": 13542}}, {"node": {"oid": "a5bef6886bd990800d25f2261fab047302bd3bed", "messageHeadline": "Threading support for Hybrid core architecture (#6728)", "author": {"name": "Ramakrishnan Sivakumar", "email": "ramakrishnan.sivakumar@intel.com", "user": {"login": "ramkrishna0116"}}, "additions": 45, "deletions": 10}}, {"node": {"oid": "6810d98ea3bb0897186f9c98a05be2961f0dad25", "messageHeadline": "Update links to gh-pages for ORT minimal documents (#6721)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 27, "deletions": 27}}, {"node": {"oid": "af4e5c0c6ee240693b7fa084c20388c89ff670a4", "messageHeadline": "Minor WinML model test skip name change", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b41e9b5d4c5b30a750aa243f94493f60d174c2d3", "messageHeadline": "[OpenVINO-EP] Fixes OpenVINO-EP build on windows (#6726)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "9d4b730e46b9e02e70d889671d92ca8748ee3207", "messageHeadline": "Fix merge leftover", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "9853ef84f871f5c15ede148d1f92ef584966a23a", "messageHeadline": "Reduce binary size, limit  asynchronous/backgroud thread stuff to tra\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 31, "deletions": 4}}, {"node": {"oid": "5b7e7aaa455019f77cf5f0a4da2a9ad8d84a440f", "messageHeadline": "Move event_pool and message_queue to core.", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "eecce31a8baf0b523d002c2c2ad8af66c86b5c5d", "messageHeadline": "Fix build, cleanup.", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 111, "deletions": 229}}, {"node": {"oid": "3184c47ad1452d4eef7092c718847c391b825667", "messageHeadline": "Merge branch 'master' into thiagofc/merge-from-master", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 23301, "deletions": 13362}}, {"node": {"oid": "14c9095d1151f2196b8eaa5d3c4be071d97be917", "messageHeadline": "Merged PR 5694922: Onnxruntime GitHub to DmlDev RI", "author": {"name": "Brian Martin", "email": "brianma@microsoft.com", "user": {"login": "martinb35"}}, "additions": 8338, "deletions": 3286}}, {"node": {"oid": "c31d27882669443d85ad4ef8bbd4a2f275fea82a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into dmldev_temp", "author": {"name": "Brian Martin", "email": "brianma@microsoft.com", "user": {"login": "martinb35"}}, "additions": 8338, "deletions": 3286}}, {"node": {"oid": "9a9202a218c085aabe735b8fbfceb92d87f99e5f", "messageHeadline": "[Node.js binding] update dependency typedoc (#6720)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 24, "deletions": 92}}, {"node": {"oid": "0be5475de6464c38c8fcc5093abf024bfd38f2ed", "messageHeadline": "Update packaging pipelines(#6664)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 53, "deletions": 55}}, {"node": {"oid": "46c06f6ac783529f4d24538524dd41897dc7c994", "messageHeadline": "Change Windows GPU CI pipeline to CUDA11 (#6616)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 17}}, {"node": {"oid": "eefeacd8288157bd03e7e8ab7d1276bb6a66dc1c", "messageHeadline": "Skip running gpt2 model in C# x86 (#6722)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b8d5fa812c35395fcb094c7ed64dc305b8a8a874", "messageHeadline": "Fix typo in README.md (#6713)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9e67b88c8312e124c3127c6ee0833c110a18fd5a", "messageHeadline": "Use local rank as GPU ID (#6719)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 4, "deletions": 13}}, {"node": {"oid": "9043df8b66472c457764816441c8bbcf3f15a7e8", "messageHeadline": "Deprecate OMP from nuget pipeline (release:1.7) (#6671)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 229, "deletions": 226}}, {"node": {"oid": "105883f4b8a88ee70f2e0bd621ff4be00a219634", "messageHeadline": "remove longformer_global_impl.cu from hipify (#6716)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 2, "deletions": 0}}, {"node": {"oid": "aa2622efb288bd5b0d09b048692899c1b5ae27cd", "messageHeadline": "Support multiple dynamic inputs in custom ops (#6666)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 181, "deletions": 20}}, {"node": {"oid": "01dfa8e125446477679abbcd3b4dc51427f69923", "messageHeadline": "Support non tuple return values from torch.nn.module (#6660)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 149, "deletions": 20}}, {"node": {"oid": "02c7873b0e2881a0302a9c1ffb09379c6e367825", "messageHeadline": "Update ORT model conversion script to support custom ops (#6701)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 110, "deletions": 75}}, {"node": {"oid": "7f33671ade046fce413dcbe18f65436e7705dbee", "messageHeadline": "Handle multiple devices scenarios (#6672)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 173, "deletions": 62}}, {"node": {"oid": "7ee5baa60db7ecfbfe956b0c8d58f1c2461f6972", "messageHeadline": "Remove monkey patch for PyTorch Nightly + ORTTrainer (#6659)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 3, "deletions": 32}}, {"node": {"oid": "9b446d5f7e51780619c95b12e12e8ec445948628", "messageHeadline": "Longformer Attention CUDA kernel memory Improvements (#6646)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 831, "deletions": 492}}, {"node": {"oid": "b09a370218c357b70b8b75f524fff3134697dbca", "messageHeadline": "Address warning in data_types_internal.h (#6704)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 12, "deletions": 9}}, {"node": {"oid": "c36ee4bd408f0c274d9ea53bd059579dcb952676", "messageHeadline": "Rename Python packaging pipelines (#6682)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "497eef8d3da5473d14c5fdc8476492544198a52d", "messageHeadline": "remove omp (#6675)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d48a4c0a541e403c697f0b7927f8f860981ca671", "messageHeadline": "Add CG step to nuget GPU pipeline (#6678)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "9a011740377bb57bc051edb0a4e3a66053c58006", "messageHeadline": "Disable some unit tests for training (#6699)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 35, "deletions": 1}}, {"node": {"oid": "33279250b57764bd65487ea91a4f92d576e03c20", "messageHeadline": "Update a couple of usages of args.minimal_build to check for not spec\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 34, "deletions": 62}}, {"node": {"oid": "37b83acd765a8bc531c062125c3ffad3b870ad82", "messageHeadline": "MLAS: add uint8_t NHWC max pooling (#6684)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 667, "deletions": 13}}, {"node": {"oid": "a35b30e2379352540b8611429c9ef83ee50faa0e", "messageHeadline": "Change BuildKernelDefConstraintsFunctorFromTypeList struct to BuildKe\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 43, "deletions": 42}}, {"node": {"oid": "f649f917fe248a388d075f47903929d3dc4547ae", "messageHeadline": "[OpenVINO-EP] Enabling OpenVINO Runtime options for Perftest applicat\u2026", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 65, "deletions": 2}}, {"node": {"oid": "df3d6bad5fccee4c161a1c68338dd3d97ccf8243", "messageHeadline": "Deprecate OMP from Python package (#6610)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 64, "deletions": 6}}, {"node": {"oid": "72eb5de0e21a140371ab07222c0f43faff94e927", "messageHeadline": "Add Python 3.9 to pypi metadata", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "25f7c935049a582f3b3b7a35da82773ef9bc7baa", "messageHeadline": "Require explicit inclusion of custom op support in a minimal build (#\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 88, "deletions": 43}}, {"node": {"oid": "dd50c39ac6bb52e09890d6a4398318dc348cd657", "messageHeadline": "Change Linux python packaging pipeline compile flags (#6668)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "87cb6fd495c046dac88893818478bb027969d611", "messageHeadline": "Add LearningModelBuilder to WinML Experimental Namespace along with v\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 3932, "deletions": 369}}, {"node": {"oid": "ff465483b10a6d895d0aa4c9795946f9c197ebe0", "messageHeadline": "Add TNLRv3 fp16 pattern to Layer Norm fusion (#6661)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 89, "deletions": 10}}, {"node": {"oid": "a07a14dce5c5a13018060080298377f9186704e1", "messageHeadline": "exclude non support types (#6653)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "b2cddc5337d4e4c058b2f57fe4a65e539c3c491b", "messageHeadline": "Consolidate MLTypeCallDispatcher classes (#6651)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 274, "deletions": 273}}, {"node": {"oid": "e6de0eb813111ff0823f99a183a7284ae8852ebc", "messageHeadline": "Add nightly pipeline for MI100 to run convergence and batch size test\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 132, "deletions": 5}}, {"node": {"oid": "f11b5d3072cc7e912d80ae7dbbb672948280be0a", "messageHeadline": "[CoreML EP] Enable coreml for onnx_test_runner and onnxruntime_perf_t\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 33, "deletions": 11}}, {"node": {"oid": "78e408dbe992a0643b56d51461cbb2c5738c53f5", "messageHeadline": "Enable type reduction for ConstantOfShape CPU kernel. (#6594)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 223, "deletions": 188}}, {"node": {"oid": "950c941f116d43a20adb5272a5ff875064d48480", "messageHeadline": "Remove year from license (#6658)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ce01c3760fc79a6fdaa42edff950bdb8b527b56d", "messageHeadline": "Cleanup macros used to register activations. (#6628)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 31, "deletions": 43}}, {"node": {"oid": "1916e35beab5be39875b426fd0ef82ac3cf050be", "messageHeadline": "Reduce tensorprotoutils binary size (#6634)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 346, "deletions": 225}}, {"node": {"oid": "fba46a76bcbe0bc5daee20f56d3a81fec546e9ef", "messageHeadline": "Update readme to reference docs webpage content (#6621)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 26, "deletions": 217}}, {"node": {"oid": "8378a45ae7fce44cc69100d17f95557ed82b10a1", "messageHeadline": "Add python 3.8/3.9 support for Windows GPU and Linux ARM64 (#6615)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 306, "deletions": 849}}, {"node": {"oid": "1c3168c0f6bdbad084a3c4fd7b34b4a6215cf646", "messageHeadline": "Skip constant folding dequantizelinear for quant qdq format (#6643)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 89, "deletions": 14}}, {"node": {"oid": "0732d72706f9d268388492dab68f122ad8bb3c79", "messageHeadline": "Add support for dynamic axes for outputs + check model output type be\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 231, "deletions": 63}}, {"node": {"oid": "b4b829dfcfc46a29780a0be54facded94ff45ce5", "messageHeadline": "Update transformers tool based on latest transformers (#6641)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 51, "deletions": 25}}, {"node": {"oid": "9294dde1431b912086ed3193d95861c9b80bd252", "messageHeadline": "Rename ONNX graphs variables in ORTModule (#6645)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 64, "deletions": 64}}, {"node": {"oid": "a7b6fc08f214e8ae0ff4c0a1ee639bfd1efdbc7b", "messageHeadline": "Support skiplayernorm fusion without beta in layernorm (#6617)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 20, "deletions": 2}}, {"node": {"oid": "fd83e38dcf81318848b07c163cbb7d37b77df90c", "messageHeadline": "[CoreML EP] Add support of BatchNormalization/Reshape/Global[Average/\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 376, "deletions": 17}}, {"node": {"oid": "64edcad2d8cd28236f8bba3a57d642d53352e652", "messageHeadline": "[NNAPI EP] Add EP option to disable CPU (#6593)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 70, "deletions": 4}}, {"node": {"oid": "d2ce8a2c800448613495ccfc196a724f72a9c081", "messageHeadline": "Add hipFFT include directory (transitional step) before ROCm. (#5992)", "author": {"name": "Matthew Emmett", "email": "matthew.emmett@amd.com", "user": {"login": "memmett"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "a9245f59d424411ae43c9e3c8d87363647d19f07", "messageHeadline": "Merged PR 5673432: Unsquash RI", "author": {"name": "Ori Levari", "email": "Ori.Levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "042964f6330a209dbc1fceb2151379181085af64", "messageHeadline": "Change how ONNX get installed", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 17, "deletions": 17}}, {"node": {"oid": "986c7a9bc033b7b0f8fa6cd0a17541b2e402a4bf", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/orilevari/un\u2026", "author": {"name": "Ori Levari", "email": "orlevari@microsoft.com", "user": null}, "additions": 2058, "deletions": 1228}}, {"node": {"oid": "e1b0cf9deea3bf1f12adeef577b724031235a351", "messageHeadline": "Revert \"Merged PR 5672827: RI from github\"", "author": {"name": "Ori Levari", "email": "orlevari@microsoft.com", "user": null}, "additions": 1228, "deletions": 2058}}, {"node": {"oid": "c6c0278872206f8e5612b76b518364d45094d411", "messageHeadline": "Merged PR 5672827: RI from github", "author": {"name": "Ori Levari", "email": "Ori.Levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 2058, "deletions": 1228}}, {"node": {"oid": "e59cb9455e4a4738293366bba90c656beea0ea60", "messageHeadline": "Add CI build with type reduction enabled (#6622)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 121, "deletions": 32}}, {"node": {"oid": "eec602e48a7d9c0140637f4ce2b9ce78d60ccfc9", "messageHeadline": "OrtModule v0.21 (#6395)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 753, "deletions": 508}}, {"node": {"oid": "352e8cb8a857e0930415c0a7f3d47f6196ffe472", "messageHeadline": "Move ORT_ENFORCE()'s within MLTypeCallDispatcher to helper class func\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 12, "deletions": 6}}, {"node": {"oid": "eef9a7a8a9a93701b490cb404e7d0ce016916634", "messageHeadline": "Update DirectML 1.4.0 to 1.4.1 for ORT 1.7 (#6636)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "85025731252c6cdfc5a7f91bba7e5ea5d2731a59", "messageHeadline": "fix CheckLearningModelPixelRange (#6632)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "88d48063fae7c29156fcfb6e035294920e0a2e19", "messageHeadline": "Log warning when GetGradientForOp() silently fails. (#6586)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 5, "deletions": 1}}, {"node": {"oid": "b09bfc86110ffe75dad52b7d6f01d6f2ade873d6", "messageHeadline": "Revert \"Remove abs in LpPool (#6303)\"", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 20}}, {"node": {"oid": "8972621138a7fd59c95c2fe9645bc959fc5c4ede", "messageHeadline": "Generate shape-independent graph if any input dimension < 2 (#6581)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "8f0b877a1d0f96dc3b48d18119ea2d839420dabc", "messageHeadline": "Enable running some ops on CUDA (#6572)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 172, "deletions": 30}}, {"node": {"oid": "505c1f30b59f662064bb252cd81aa8719a5c7b90", "messageHeadline": "use == instead of is for python 3.8", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e70344e648501dd0888a56fce88d43cc20fc1990", "messageHeadline": "Fix training python packaging pipeline (#6613)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "1c727742327facc89a13ecd91d8dd22770146355", "messageHeadline": "Update a few WinML model test filters for DML", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "8f14b8bd9d0706caea8ca676ecb3131bd087cd8e", "messageHeadline": "Support disabling training kernels as part of a reduced build (#6557)", "author": {"name": "Cian Hayes", "email": "cianh@microsoft.com", "user": null}, "additions": 4, "deletions": 1}}, {"node": {"oid": "e9d03983fc63828f84a9dcc8538392e1b22b46bb", "messageHeadline": "Add engine decryption in TensorRT EP (#6612)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 93, "deletions": 6}}, {"node": {"oid": "0b89f931d07860ca26ddbb9b25d70dd721e371dd", "messageHeadline": "Update CUDA build configs (#6598)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 28}}, {"node": {"oid": "d3a2c8c1c7761e000c660c3917b16e50d16ed251", "messageHeadline": "Support double for operators ReduceMax, ReduceMin (#6265)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 169, "deletions": 62}}, {"node": {"oid": "ff063309b083e44ad697caf42bd45c8d98782e0f", "messageHeadline": "enable omp for debug build", "author": {"name": "Randy Shuai", "email": "rashuai@microsoft.com", "user": {"login": "RandyShuai"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "6c5f50d00ecf8b2c06de7a85e4a92d14c9283141", "messageHeadline": "deprecate omp in ci", "author": {"name": "Randy Shuai", "email": "rashuai@microsoft.com", "user": {"login": "RandyShuai"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "56e4e47f66caecf54a5b2882cd99f566b78f248d", "messageHeadline": "Quantize model with QDQ format (#6541)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1281, "deletions": 897}}, {"node": {"oid": "c02ae61cab5a6652cb8d3854ae3ba4cec38f49a6", "messageHeadline": "Make kernel hash stable in type reduced build (#6603)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 255, "deletions": 130}}, {"node": {"oid": "16eed68a1eea02799a8879a148f1fb17ede59f63", "messageHeadline": "Fix layer_norm.cc on x86 (#6556)", "author": {"name": "Cian Hayes", "email": "cianh@microsoft.com", "user": null}, "additions": 8, "deletions": 4}}, {"node": {"oid": "97c1693305927438c3859a9155de1bab203faded", "messageHeadline": "Merged PR 5663009: Merge latest github onnxruntime into dmldev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 10778, "deletions": 7506}}, {"node": {"oid": "284f68638fcfd4f3fbb464a7a3582f394a2daf87", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 10778, "deletions": 7506}}, {"node": {"oid": "13d7db9a98542eb40399592684309ff7dd259551", "messageHeadline": "Don't update the excluded ops/types unless args.update is true. Updat\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "0b1e21c638d5ba434527b29b8df518bc34304d14", "messageHeadline": "Fix bug with ORT format serialization of tensor attributes. (#6602)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 22, "deletions": 17}}, {"node": {"oid": "67ef6b1aa63ae518b98da27086fa312cf0442c57", "messageHeadline": "[Mult-GPU inferencing] Add new API to get/set device id. Set correct \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 72, "deletions": 0}}, {"node": {"oid": "1dd920fa7cac2d2ce8a0b497f2f32bd4a2b099fc", "messageHeadline": "Fix TensorRT unnecessary file cache operations (#6601)", "author": {"name": "nietras", "email": "nietras@users.noreply.github.com", "user": {"login": "nietras"}}, "additions": 29, "deletions": 27}}, {"node": {"oid": "19c130f561de5144a0c97d71373d7232938a202e", "messageHeadline": "Reduce CastMLFloat16ThroughFloat size (Scott's suggested changes), fi\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 48, "deletions": 48}}, {"node": {"oid": "190b90a6829b458c32533790ea30cd6238b71d4b", "messageHeadline": "Fix some coding conventions issues (#6583)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 24}}, {"node": {"oid": "c86c21e002115ded3aad590b0abaf02688c2f446", "messageHeadline": "Generate error when an explicit stream argument is not provided in th\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "d18aa45b4622c569a1ecaf601b5d0514f5fea645", "messageHeadline": "Enable more ROCM ops that are sharing CUDA code.  Some are needed for\u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 30, "deletions": 47}}, {"node": {"oid": "dbe31361bc111e83fbf1a8c9560f2b8394ae8938", "messageHeadline": "Fix build.gradle so it always targets Java 8 class files.", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 17, "deletions": 0}}, {"node": {"oid": "b57a7f4de3e60557d8774a5803fc40bd55c23180", "messageHeadline": "Delay load dxcore in winml model tests", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 3, "deletions": 0}}, {"node": {"oid": "b50b0a89aac7d464c55978861213151efa8f4494", "messageHeadline": "Fix build failure when building with --build_wheel on Windows", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "af9dfa7a4d3c2b83bdccc57c76e3e502cbcc70ae", "messageHeadline": "Remove docs that have been migrated to https://onnxruntime.ai/docs (#\u2026", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 5, "deletions": 3022}}, {"node": {"oid": "dda5a6207275b330c1571d13f27a3270dcf6e54f", "messageHeadline": "Fix updated Doxygen errors. (#6588)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 19, "deletions": 0}}, {"node": {"oid": "115e16b37b6f5a57dd8a39432b78d6e7e2c63293", "messageHeadline": "ort_test_utils: skip creating input if it is an initializer (#6544)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "ccfd90291bda29492b420037ad57237b0470798b", "messageHeadline": "Remove condition from ORT_RETURN_IF[_NOT] macro output. (#6563)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 156, "deletions": 151}}, {"node": {"oid": "b5bd14fc9fddb9df41405b0661a39c026f7316db", "messageHeadline": "Update GPU packaging pipelines to cuda11 and fix the other build brea\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 52, "deletions": 53}}, {"node": {"oid": "82229c8e61062ef17e57b9ca4f1189f3b589cfcb", "messageHeadline": "Support no bias in layernorm and skiplayernorm op (#6554)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 122, "deletions": 35}}, {"node": {"oid": "299ace0759192557e97e14d03c08f0334c84e9d5", "messageHeadline": "Support to allow user to specify compute stream per session (#3723)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 1877, "deletions": 1110}}, {"node": {"oid": "973c3917a68036c6627fd908f106b68153271596", "messageHeadline": "OpenVino add build_shared_lib flag in the build command (#6560)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "68193e28dedfb0d3847d924c550a4d74646e2bed", "messageHeadline": "Let execution fall back to CPU EP if Compile of a partition on curren\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 162, "deletions": 34}}, {"node": {"oid": "f2ce3aae1398490e483f1853d6d41dda4f856c8f", "messageHeadline": "add set_model_dir and update ONNX (#6119)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "3b376da37c0d720991ea45b68a95e0b0853b9037", "messageHeadline": "Enable type reduction for Gather CPU kernel. (#6579)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 29, "deletions": 15}}, {"node": {"oid": "c5d2538314d17e2351a0f002ce8a70fede435930", "messageHeadline": "Add more kernels that have typed registrations to the operators we tr\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 23, "deletions": 9}}, {"node": {"oid": "f14c621c103e1880aa12abd5d148c98318741f5f", "messageHeadline": "Tile perf enhancements - continued (#6561)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 211, "deletions": 39}}, {"node": {"oid": "c49d1dbc4b71cd3e0f3531504c0c0d805aa45cd2", "messageHeadline": "Add type reduction support to Slice and Transpose (#6547)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 275, "deletions": 154}}, {"node": {"oid": "89627a81787c514ca9af4022857c330327408b94", "messageHeadline": "[Node.js binding] support NPM v7+ (#6559)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 6336, "deletions": 818}}, {"node": {"oid": "615acf156c87f809e3aac635fbde56af12de2e90", "messageHeadline": "remove keras example from python documentation (#6574)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 2, "deletions": 96}}, {"node": {"oid": "4e61e254ecf3c6de5388efb3678208c127d3104e", "messageHeadline": "Update link in readme (#6537)", "author": {"name": "Prasanth Pulavarthi", "email": "prasantp@microsoft.com", "user": {"login": "prasanthpul"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d914e29fe1f165ce7055ae402306f72473a5fd59", "messageHeadline": "Reuse reduction_functions.cu", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 0, "deletions": 480}}, {"node": {"oid": "3c441849635cec342da5906720742a6c9632a41a", "messageHeadline": "Pick up changes from:", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 22, "deletions": 16}}, {"node": {"oid": "a9e4d70b508becf51430bde292fc67533cd3e848", "messageHeadline": "Fix merge conflict.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "76fcebd0a49ad6e93b877061c93cbabbf3f41d4c", "messageHeadline": "Fix scratch buffer early free.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 14, "deletions": 6}}, {"node": {"oid": "86ac11af1a04bd5c60e7b28e725ea66a9dadcc7c", "messageHeadline": "Delete ROCM-specific reduction code that is identical to CUDA reducti\u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 13, "deletions": 863}}, {"node": {"oid": "5d8792705b062b8386fc58470f509f9696ab0380", "messageHeadline": "Code formatting.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 79, "deletions": 80}}, {"node": {"oid": "21a47ec8d985f8fd723a2baff842315f1b2d9a56", "messageHeadline": "Disable a couple more unsupported tests.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "0b147702af9de6ddede6f139223454b545baacd9", "messageHeadline": "Update remaining reduction ops to use MIOpen.  double datatype is not\u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 745, "deletions": 788}}, {"node": {"oid": "a28ddb85b63c465388fa60faab15fa778a20ff60", "messageHeadline": "Reduction ops.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 114, "deletions": 14}}, {"node": {"oid": "196132925ece5590acb6d39510ae4b19009ed221", "messageHeadline": "Reuse CUDA's reduction_functions.cc", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 51, "deletions": 4}}, {"node": {"oid": "4c1db50df52c889bccc538a8dd83afec4e4bb095", "messageHeadline": "miopen common", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 71, "deletions": 21}}, {"node": {"oid": "554184bcc493ccd6b26d3b6666d898d8d5243947", "messageHeadline": "Add reduce template parameters.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 31, "deletions": 30}}, {"node": {"oid": "c4b6559be93621f919086788fdcd9969a841d69b", "messageHeadline": "Update reduction_all.cu", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "5fc377f21e7988b2d392a489393aeccab79d4d1c", "messageHeadline": "Partial updating of ROCM reduction code.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 698, "deletions": 26}}, {"node": {"oid": "318b82ca7e6e73ed46fc4d09e97466181ef35a9f", "messageHeadline": "Cast Op performance fix. (#6509)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 345, "deletions": 393}}, {"node": {"oid": "7427d6e10c9e0e96a869b3080f04e3ab3a5939b5", "messageHeadline": "Merged PR 5650614: Merge Github onnxruntime into dmldev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 31381, "deletions": 11322}}, {"node": {"oid": "355cc0e62e09fbec92486221b9879d136b2cf809", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/rylai/update\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 31381, "deletions": 11322}}, {"node": {"oid": "2ef792ae6eb9e7e0afc260d0bc1d73a26edc6914", "messageHeadline": "Don't resolve symlink in resolve_executable_path(). (#6540)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "aa31ba57745b6196827d20f8b817b6053332e1ed", "messageHeadline": "Merge CPU packaging pipelines (#6480)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1925, "deletions": 1992}}, {"node": {"oid": "bc0d04bf07c7fdab1cb75c03f3b626d6f31d70a7", "messageHeadline": "Revert \"Add support for dynamic axes for outputs + check model output\u2026", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 54, "deletions": 218}}, {"node": {"oid": "0d35f0e2c05ed003f22b1383223a9604f594ea33", "messageHeadline": "[CoreML EP] Add support of Conv operator (#6510)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 278, "deletions": 25}}, {"node": {"oid": "6cf54ff296eb79b31be3f061499b907100e18c5f", "messageHeadline": "Switch Android CI java build to JDK 11  (#6552)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 20, "deletions": 1}}, {"node": {"oid": "c7feb48083d0445794365d70a15a4587179cf541", "messageHeadline": "Don't send out Runtime error telemetry when can't create LearningMode\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 24, "deletions": 16}}, {"node": {"oid": "464dbef143fdf272edba26d09b596bc85b758659", "messageHeadline": "[NNAPI EP] add uint8 support for Transpose/Concat/Maxpool, add suppor\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 498, "deletions": 68}}, {"node": {"oid": "c983b8431655dfb1c204d735a5c16c31f52e92fd", "messageHeadline": "Add support for dynamic axes for outputs + check model output type be\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 218, "deletions": 54}}, {"node": {"oid": "62ac16427931345f3c67b4c9fc747f31b58f5a23", "messageHeadline": "Cache datasets on CI machines (#6525)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 112, "deletions": 51}}, {"node": {"oid": "6cb8f8c812fc0d793976e94ab4275384337430fd", "messageHeadline": "Support disabling a typed kernel registration that uses the output ty\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 30, "deletions": 18}}, {"node": {"oid": "8d53ef69e503faabd0f180c95e18a73b8885d059", "messageHeadline": "Add type reduction support to Min, Max and Pow (#6519)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 188, "deletions": 56}}, {"node": {"oid": "fbb24b57d04589316262f49cd69e64a3027ee81c", "messageHeadline": "Update code owners for pytorch frontend team (#6329)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 10, "deletions": 11}}, {"node": {"oid": "85434273ffe76b5a4d941d8a2c628df671333e48", "messageHeadline": "Fix CUDA Reduction kernel for ArgMax/ArgMix for when reduction dim=1 \u2026", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 36, "deletions": 17}}, {"node": {"oid": "8a890ddfd749a2f46609f8e234f192780bb41356", "messageHeadline": "Sync ORTModule branch with master and fix tests (#6526)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 45548, "deletions": 17197}}, {"node": {"oid": "14f7d56c81beb49796416b5f8aea7e41c061f870", "messageHeadline": "Add optimized version of ConvGrad for pointwise convolutions. (#6531)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 156, "deletions": 49}}, {"node": {"oid": "6fc5237d9e79dac76886368d4b5c7e3d81d5a8e2", "messageHeadline": "Introduce --enable_training_ops build flag (#6523)", "author": {"name": "Cian Hayes", "email": "cian.m.hayes@gmail.com", "user": {"login": "cianmhayes"}}, "additions": 47, "deletions": 9}}, {"node": {"oid": "9a6e71574a2cc85a784eaef497b2e158429316d9", "messageHeadline": "MLAS: improve quantized depthwise convolution (#6513)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 835, "deletions": 373}}, {"node": {"oid": "588ddeb82fb3719ee2b829ebdc7705c6db93f9ed", "messageHeadline": "Add level 1 optimized mnist model so that the required_ops.config inc\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "7264a067a9116289a3a4e731d305973fc96f0369", "messageHeadline": "Implement QuantizeLinear with avx512 (#6260)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 372, "deletions": 5}}, {"node": {"oid": "5b69cbe80e170bd90349bbc603304571c28b8af3", "messageHeadline": "Fix Windows CI builds by updating test scripts to work with numpy 1.2\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 21, "deletions": 18}}, {"node": {"oid": "e5cbcec17f839076e05b1435c9b2769b3e9dda65", "messageHeadline": "Fix issues with ArmNN build setup (#6495)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 22, "deletions": 28}}, {"node": {"oid": "f2872ffd649fb0d6c0b2045736caae8bae350526", "messageHeadline": "Print a warning message for using newer c_api header on old binary (#\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "7c5bfbaaab083aa270a4ca6c54dc88f2ff2b6d1a", "messageHeadline": "Lochi/refactor yolov3 quantization (#6290)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 513, "deletions": 33}}, {"node": {"oid": "a36f627a4c20317203247239ad3ded4fda328610", "messageHeadline": "Dnnl training (#6045)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 2029, "deletions": 112}}, {"node": {"oid": "3a30ad7b5782149209e8bd661dcdd3b81fea1c39", "messageHeadline": "handle hr error conditions (#6449)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "531eb064abe09600ac964d8b984717cc9cdc36e8", "messageHeadline": "fix sdl bugs for uninitialized variables and returns (#6450)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 12, "deletions": 6}}, {"node": {"oid": "76f5d9edc6e0f165c9b6b072e87b911200e2e70d", "messageHeadline": "add explicit barriers for buffer overread and overrwrite (#6484)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 12, "deletions": 8}}, {"node": {"oid": "7f5731741dbd48cfeb560dfc82a7189b5f2083a6", "messageHeadline": "Optimize GatherGrad for AMD GPU (#6381)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 71, "deletions": 13}}, {"node": {"oid": "76bc0e479c9988152598bcffde96b7f77fbcbe1f", "messageHeadline": "Enable dense sequence optimized version of Pytorch exported BERT-L on\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 37, "deletions": 24}}, {"node": {"oid": "8c6d76a4c0c6718405c61d9eb19f9920fc32af35", "messageHeadline": "Update to match new test setup. (#6496)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 7}}, {"node": {"oid": "8306150e0e5f63fb51e0574fc695ae3cdbd8503e", "messageHeadline": "Refine transformers profiler output (#6502)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 216, "deletions": 77}}, {"node": {"oid": "06a6c634341d5b1dfe691c8ed3bd44bdda85e6bd", "messageHeadline": "[CoreML EP] Add support for some activations/Transpose, move some sha\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 382, "deletions": 140}}, {"node": {"oid": "a19c48f5cbb345d8d8355414d86410629dff0d25", "messageHeadline": "Fuse cuda conv with activation (#6351)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 713, "deletions": 298}}, {"node": {"oid": "71389ff9ab80d4c89aefe9215cc11844de3b9bd4", "messageHeadline": "nuphar test to avoid test data download to improve passing rate (#6467)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 15, "deletions": 24}}, {"node": {"oid": "d3203adc26028ce81d72115e17c06c8cc2ff85ba", "messageHeadline": "Update document of transformer optimization (#6487)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 39, "deletions": 17}}, {"node": {"oid": "066520f6c14aac1a9aa196ef1d0291cd306e5b84", "messageHeadline": "Improve work distribution for Expand operator, and sharded LoopCounte\u2026", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 59, "deletions": 42}}, {"node": {"oid": "7abb5b667fe52deafcbfe432bca56e60509312a4", "messageHeadline": "Support pad operator in quantization and quantized nhwc transformer. \u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 632, "deletions": 18}}, {"node": {"oid": "1a5b75a55418c9478bbe4d4ad28ffa57c3f5a924", "messageHeadline": "[OpenVINO-EP] Remove support for OpenVINO 2020.2 (#6493)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 59, "deletions": 293}}, {"node": {"oid": "3b1227c5ce7866e5879a7872bc6a3b45f1871646", "messageHeadline": "SDL annotation fixes (#6448)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 64, "deletions": 63}}, {"node": {"oid": "21b4842c3457e159089f1df1c99eca741035818c", "messageHeadline": "SDL fixes: add proper casts/format specifiers (#6446)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "d4e1f5ab783e1c8d5ab821f5e4d1f860f066fe5b", "messageHeadline": "Fix of support api version bug for [de]quantize (#6492)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "785e51d22f6613218cd1984a7785804f5333f0c8", "messageHeadline": "Export the model with torch.no_grad() context (#6472)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 69, "deletions": 9}}, {"node": {"oid": "ce46f37ff21291e49c67d985794661ec563a93c9", "messageHeadline": "expose learningmodelpixelrange property (#5877)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 84, "deletions": 50}}, {"node": {"oid": "93aa72e468f1c71b9a26065c7e1a8870b54e21f9", "messageHeadline": "Support inputs to ORTModule forward method that require gradient (#6420)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 95, "deletions": 25}}, {"node": {"oid": "3f60b27703d40fcbedf62e8413f409f304f115fb", "messageHeadline": "Speed up the Mac CI runs (#6483)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "ea2b5600557860e39800b5caa69bc8ea5f641b56", "messageHeadline": "Fix test breaks in Windows ingestion pipeline (#6476)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 37, "deletions": 58}}, {"node": {"oid": "00afd00059a40e56ea96f0d6b7d96c1848b077c7", "messageHeadline": "merge e2e with distributed pipeline (#6443)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 19, "deletions": 68}}, {"node": {"oid": "c84bb9df9f01cbd31ceaebcbf0d30f2e327c16a1", "messageHeadline": "Add ability to track per operator types in reduced build config. (#6428)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3292, "deletions": 438}}, {"node": {"oid": "752627c5bb72eabbab73425199d83c4004ba23ab", "messageHeadline": "[CoreML EP] Add CI for CoreML EP (macOS) and add coreml_flags for EP \u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 111, "deletions": 39}}, {"node": {"oid": "2e228d74d07a10f7a85d7c127dd30f36e2bae30b", "messageHeadline": "Increase the distributes tests pipeline timeout to 120 minutes (#6479)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "77d0eb3f56fcd91868170decbbf36f8a9e157177", "messageHeadline": "Fixing a leak in OnnxSequences with String keys or values. (#6473)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "237b275bd8883c99de7a1ba4160d81c7b6ac7df9", "messageHeadline": "Enable device change during training + minor forward() refactoring (#\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 82, "deletions": 57}}, {"node": {"oid": "d850fa63bf59177a2f03e7e6ad3faa7d7b37e2ea", "messageHeadline": "Op kernel type reduction infrastructure. (#6466)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 750, "deletions": 380}}, {"node": {"oid": "91b19b836451db2f0faa1733f2f14ba3408b1ada", "messageHeadline": "Delete nuget extra configs (#6477)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 16}}, {"node": {"oid": "7a0ab9c450fff3d24276e828dd887036f9c829f1", "messageHeadline": "Update pypi package metadata (#6354)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 13, "deletions": 5}}, {"node": {"oid": "b6ac35fed3988737a2af5e73a156a097be2cbdb5", "messageHeadline": "use tickcount64 (#6447)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "ed1ebd2e2109aca68f678fbefa6a36caa77429cf", "messageHeadline": "fix SDL rule (#6464)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "1ce1a51d4622775d29a619c194037b8dbd14df87", "messageHeadline": "fix SDL native rule warning #6246 (#6461)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0100f336d7547486432e6808256d0b4938581a1b", "messageHeadline": "[java] Adds support for OrtEnvironment thread pools (#6406)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 387, "deletions": 34}}, {"node": {"oid": "f68eb35aed1a68f0c98aacebc40f593f99f12ac5", "messageHeadline": "dequantize 1st input of lstm back if it is quantized (#6444)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "d5f51c40336b55f2e9cd27bb7585c1627df3d133", "messageHeadline": "Bug 31463811: Servicing: Redist (Nuget) conflicts with Microsoft.AI.M\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 34, "deletions": 18}}, {"node": {"oid": "c05adb114713443a5d3e9398592aa2ea36b86161", "messageHeadline": "Initial version of CoreML EP (#6392)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1883, "deletions": 6}}, {"node": {"oid": "fd43806252ecd5a72a90b84c25fbce0bc21506b6", "messageHeadline": "fix max norm clipping test in python packaging pipeline test (#6468)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 56, "deletions": 24}}, {"node": {"oid": "b5d1a49b300e2e2b2182f90b21b0dd749663b337", "messageHeadline": "Share allocator between CUDA EP & TRT EP. (#6332)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 281, "deletions": 111}}, {"node": {"oid": "9835b46a1df5c1dfeb634c61668c644390577d47", "messageHeadline": "Add an option to save the training graph after optimization (#6410)", "author": {"name": "Ryota Tomioka", "email": "ryoto@microsoft.com", "user": {"login": "ryotatomioka"}}, "additions": 38, "deletions": 13}}, {"node": {"oid": "0d20104a7251fe7392ea44874129762127c4731c", "messageHeadline": "only build experimental api in redist (#6465)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 34, "deletions": 13}}, {"node": {"oid": "afd7b8b3f751e7aedbea708282003c380084606a", "messageHeadline": "add tool for generating test data for longformer (#6415)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 241, "deletions": 0}}, {"node": {"oid": "76dbd88526dcb177a7b73ac0476e030b695952dc", "messageHeadline": "Expose graph ModelPath to TensorRT shared library (#6353)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 43, "deletions": 24}}, {"node": {"oid": "7e42840298aa09b35a709bd327edca110691d6eb", "messageHeadline": "fix null dereference warning (#6437)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 43, "deletions": 48}}, {"node": {"oid": "f3a0344f9a9c8256d114ae9b930eaaa0c1f37592", "messageHeadline": "Farewell TrainableDropout (#5793)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 116, "deletions": 344}}, {"node": {"oid": "6ed12402a471796cf190db230e1cbbfd970efdbf", "messageHeadline": "Liqun/liqun/enable pipeline parallel test2 (#6399)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 50, "deletions": 164}}, {"node": {"oid": "24f1bd6156cf5968bbc76dfb0e801a9b9c56b9fc", "messageHeadline": "Minor cmake change (#6431)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 0, "deletions": 4}}, {"node": {"oid": "c20965f9b2a241bf3c675f7f4dabb4890e5f3296", "messageHeadline": "enable pipeline to run quantization tests (#6416)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 24, "deletions": 14}}, {"node": {"oid": "e1dc268e45f1f37a7b46d37e47d65416832b1db7", "messageHeadline": "Add support for custom ops to minimal build. (#6228)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 465, "deletions": 211}}, {"node": {"oid": "6507b4f818455671007b10d27b8d59f7f36f223f", "messageHeadline": "Reintroduce experimental api changes and fix remote build break (#6385)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 181, "deletions": 16}}, {"node": {"oid": "3c3d36334b051cf809f646ecfc42677edae3ea11", "messageHeadline": "Continue memory planning when unknown shape tensor is encountered. (#\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "61ecf52c2438a2192c73c16032f3c62a4cfbe65a", "messageHeadline": "Fix generate_submodule_cgmanifest.py Windows issues. (#6404)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "60c772e2bc3a48af5621c895558038e764a51fd3", "messageHeadline": "Megatron checkpointing (#6293)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 3576, "deletions": 228}}, {"node": {"oid": "4442d94c6cf2a144a6b08336d35a603a6efb4b97", "messageHeadline": "OpenVino docker file changes to bypass privileged mode", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 64, "deletions": 17}}, {"node": {"oid": "bba185a582655b95fdd9913ac45f3f5c8ffd10ba", "messageHeadline": "Fix some compile warnings (#6316)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 345, "deletions": 322}}, {"node": {"oid": "98cc7b5a9eaee2f9ed44e26c05d65c5693e32c4a", "messageHeadline": "Load the model path correctly (#6369)", "author": {"name": "Martin Man", "email": "supermt@gmail.com", "user": {"login": "MartinMoon"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "99a38f40232719ad2da39f06cb469c11c971353c", "messageHeadline": "fix build on cuda11 (#6394)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "eb946c4177ccc6d41ed4dae0f3f2500ebfdc0441", "messageHeadline": "Unblock Android CI code coverage failure (#6393)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8574854d23595ee7fb37037bceea72345d23ed1d", "messageHeadline": "[Perf] Optimize Tile CPU and CUDA kernels for a corner case (#6376)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 179, "deletions": 19}}, {"node": {"oid": "d9e4795385cd3bdf7391b4a02290ac954aa8063d", "messageHeadline": "Fix Windows x86 compiler warnings in the optimizers project  (#6377)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 23, "deletions": 24}}, {"node": {"oid": "33f60a06d556b0ce547212e02094839501dffb19", "messageHeadline": "Dont use default string marshalling in C# (#6219)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 46, "deletions": 25}}, {"node": {"oid": "69af0440b1e19645fb4de4b76814cfd5cb138c48", "messageHeadline": "Add the custom op project information (#6334)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "453431f7bb7dc7a577e86b9ef16b6e0a2b656627", "messageHeadline": "Add max_norm for gradient clipping. (#6289)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 649, "deletions": 233}}, {"node": {"oid": "a1b5bfc4f82beaa616b1f5d8b035494b5ccc02cf", "messageHeadline": "Fix SDL warning (#6390)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 17, "deletions": 10}}, {"node": {"oid": "d7bdd964250f45adfcf897236d3fbd6690bce932", "messageHeadline": "Refine auto_pad based pad computation in ConvTranspose (#6305)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 90, "deletions": 6}}, {"node": {"oid": "ac36596fb81ef68411f35de954588b3850b0d4d8", "messageHeadline": "fix convert_common version retrival (#6382)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "910c5ab65508c82fdd40d4b6b53ec4df7978e9dc", "messageHeadline": "Add ORTModule deepspeed zero stage 1 test to the distributed CI pipel\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 31, "deletions": 12}}, {"node": {"oid": "baac7c91e245eb5a28476622ffff9e993acd594f", "messageHeadline": "Support MLFloat16 in CumSum Cuda op for Opset 14 (#6355)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 37, "deletions": 6}}, {"node": {"oid": "5b6753ce278166bf1f979eee1643a0bec6696c40", "messageHeadline": "Wezuo/memory analysis (#5658)", "author": {"name": "wezuo", "email": "49965641+wezuo@users.noreply.github.com", "user": {"login": "wezuo"}}, "additions": 843, "deletions": 41}}, {"node": {"oid": "4db4982a5ede89a3555badc91fd9907bc61f384e", "messageHeadline": "This added telemetry isn't needed (#6363)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 0, "deletions": 14}}, {"node": {"oid": "eab164e1a5cd7760ee653ffe455e7a10cded7cb2", "messageHeadline": "Add python example of TensorRT INT8 inference on ResNet model (#6255)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 368, "deletions": 30}}, {"node": {"oid": "f5a4f7fc2aa5b89e10c05ff36aab0216d66ca09c", "messageHeadline": "fix -Wdangling-gsl (#6357)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "c8e37e3a3688c0655f6d87045d3d23cf900709f3", "messageHeadline": "Fix one more SDL warning (#6359)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "961bb62ae4f59bd36e34c9c6a1880c5b50887e0a", "messageHeadline": "Add create session to WinML telemetry to track WinML Usage (#6356)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 15, "deletions": 5}}, {"node": {"oid": "8ce252caa9e636aada6fa833329d054e68b4c9ae", "messageHeadline": "Pipeline Parallel Experimental Python API (#5815)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 2820, "deletions": 511}}, {"node": {"oid": "6d0fb3ebb3e525cd99de9b9fa57f8f069c9922ad", "messageHeadline": "Java: Set C language warnings to W4 and adjust JNI code (#6347)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 70, "deletions": 41}}, {"node": {"oid": "e54e2f969db6db06425170fb6af2b6e8884b45af", "messageHeadline": "Use readelf for minimal build binary size checks. (#6338)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 192, "deletions": 23}}, {"node": {"oid": "5d9552cc8b99d8a29e95a9edf903b24e981ea03b", "messageHeadline": "fix longformer benchmark io_binding output_buffers (#6345)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 10, "deletions": 8}}, {"node": {"oid": "ea6789b75436992c6a7c57946f3abe0b51f23847", "messageHeadline": "Add PREfast to python packaging pipeline (#6343)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 100, "deletions": 1}}, {"node": {"oid": "fd21c84eb814371ea0dd86b4902dd137846e07c9", "messageHeadline": "Enable graph save for orttrainer (#6333)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 138, "deletions": 4}}, {"node": {"oid": "c24f2950bf7a54f79f1977fd566bdc688a63953f", "messageHeadline": "update quantize to support basic optimization and e2e example for ima\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 546, "deletions": 364}}, {"node": {"oid": "5b9d993a2e76576a101755b12d9965b45f28f352", "messageHeadline": "Fix DerefNullPtr issues raised by SDLNativeRules. (#6348)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "4df356d1c9f1387897ccb9df1d170ba80e929756", "messageHeadline": "Train BERT Using BFloat16 on A100 (#6090)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 961, "deletions": 259}}, {"node": {"oid": "e35db194e3e8fad2a73e5f856fd68b75c574d45c", "messageHeadline": "fix the pipeline failure (#6346)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "042053c55e634b925ad4513d89849a1d67bc259a", "messageHeadline": "Add support for running Android emulator from build.py on Windows. (#\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 438, "deletions": 133}}, {"node": {"oid": "b220feee2ff11ba90842b38c192de8722bb4b80c", "messageHeadline": "[NNAPI] Add pow support (#6310)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 64, "deletions": 15}}, {"node": {"oid": "fcd9fc9b6d6b9b41e982a0a3ae982f6c0e54fa9b", "messageHeadline": "remove gemmlowp submodule (#6341)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 26, "deletions": 256}}, {"node": {"oid": "cfd6f100989e94b4993f12f5f9799e3e85833121", "messageHeadline": "Remove OpSchema dummy definition. Only needed for Function now, and w\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 5}}, {"node": {"oid": "d367941cc43e796e1402a0d190ddf4ef68510903", "messageHeadline": "changed wording. (#6337)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f7034b9bca705b31fa69f31486416795e0eccbd9", "messageHeadline": "add external data support to tensor proto utils (#6257)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 430, "deletions": 87}}, {"node": {"oid": "62e404591a6d76b5b40abfdddf1e355ec23bad24", "messageHeadline": "Enable add + softmax fusion for Rocm platform (#6259)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 551, "deletions": 14}}, {"node": {"oid": "56ab2166e8edcdc50455044d0d2363122c4c7147", "messageHeadline": "Delete float16.py (#6336)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 0, "deletions": 225}}, {"node": {"oid": "0586c610b2f68b4879203bc40d7367617be361ca", "messageHeadline": "Add ORTModule BERT classifier to CI the pipeline (#6330)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 37, "deletions": 16}}, {"node": {"oid": "9b7510d88c4d1166f68d85821c2c4b0769b58748", "messageHeadline": "Add ORTModule distributed CI pipeline (#6278)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 149, "deletions": 0}}, {"node": {"oid": "87ec1f6208f2bb90b079baf444ad942aecfb56ee", "messageHeadline": "MLAS: add fallback implementation for quantized GEMM (#6335)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 354, "deletions": 627}}, {"node": {"oid": "5623cc6d172b1150af2f0938455d03432605f1f4", "messageHeadline": "Use onnxruntime_USE_FULL_PROTOBUF=OFF for the cuda execution provider\u2026", "author": {"name": "Alberto Magni", "email": "49027342+alberto-magni@users.noreply.github.com", "user": {"login": "alberto-magni"}}, "additions": 0, "deletions": 5}}, {"node": {"oid": "aeca96caba284b37e27f0621f644ef2c4e8ad3b3", "messageHeadline": "Liqun/enable pipeline parallel test (#6331)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "f77ff1bc3d2f748baf4e06ba1d0bc2c089ef118a", "messageHeadline": "Quantization support for split operator with its NHWC support (#6107)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 210, "deletions": 27}}, {"node": {"oid": "6b73bae0352541059a1c464a09f57f98b3edb5d2", "messageHeadline": "Java: add Semmle to Java publishing pipelines (#6326)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 30, "deletions": 0}}, {"node": {"oid": "aacc8dbfa384186f47f5842d6c4f223b51bba469", "messageHeadline": "Remove false positive prefast warning from threadpool (#6324)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "0ed56d491ab235c4970aff181d9e6e86269431ca", "messageHeadline": "fix opset imports for function body  (#6287)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 123, "deletions": 11}}, {"node": {"oid": "b491d7c17951f74c3fcc7ab5704081288d16cb20", "messageHeadline": "Avoid false sharing on thread pool data structures (#6298)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 73, "deletions": 3}}, {"node": {"oid": "ec81e29c8459d0d113d86f7c89674b02636fa74e", "messageHeadline": "Add longformer to  python package (#6314)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 230, "deletions": 208}}, {"node": {"oid": "a8257666bd1fb4a56ed18a9a081e81cb9057ce31", "messageHeadline": "Support 1D input for Conv + Mul/Add fusion optimizer with test (#6295)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 25, "deletions": 27}}, {"node": {"oid": "3b3e698674dca2014b91fb617e2c4f22ffd0c5c9", "messageHeadline": "Remove abs in LpPool (#6303)", "author": {"name": "Luyao Ren", "email": "375833274@qq.com", "user": {"login": "luyaor"}}, "additions": 20, "deletions": 1}}, {"node": {"oid": "a038924bee0ddd3f258cadaf576fce6b7d6448d4", "messageHeadline": "update transformers required package versions (#6315)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 66, "deletions": 69}}, {"node": {"oid": "c43ca45c4fb5bf0d438d9cf8d20619faea55e1fd", "messageHeadline": "Force reinstall onnx python package on Windows (#6309)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 19, "deletions": 0}}, {"node": {"oid": "ac5b5e5d1e4f27cd7fa541cc1ca6d16e36db4949", "messageHeadline": "more dtype for Equal CUDA kernel (#6288)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 25, "deletions": 11}}, {"node": {"oid": "938e65d8780837766384736a5a6138227281f23e", "messageHeadline": "add --sequence_lengths option (#6285)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 84, "deletions": 74}}, {"node": {"oid": "f3a479901bbdfbab1564e74e3c7b88757af134ab", "messageHeadline": "add poc test for ortmodule (using MNIST dataset) to the ci pipeline (\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 19, "deletions": 6}}, {"node": {"oid": "84024bdfa98ea781ed4f7c1973425bfcb4617eb1", "messageHeadline": "Enable ONNX backend test of SequenceProto input/output  (#6043)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 16, "deletions": 13}}, {"node": {"oid": "5084ce09693439d648d6f6a844d0ea6682b16b95", "messageHeadline": "Update nuget build (#6297)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 210, "deletions": 1012}}, {"node": {"oid": "fa851bff669df92efac76dbb4ab9a3e1edc14b8c", "messageHeadline": "Add workaround to remove ROCm-specific binary-elementwise files.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 0, "deletions": 643}}, {"node": {"oid": "1059bfaf75bce27f2eeb07f6968bc307d6181924", "messageHeadline": "Workaround for static_cast<double>(half)", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 30, "deletions": 20}}, {"node": {"oid": "f9cc6ee7c29d0c28411bd395d4a321815e1deb0b", "messageHeadline": "Merged PR 5557454: Update to latest ORT master", "author": {"name": "Patrice Vignola", "email": "pavignol@microsoft.com", "user": null}, "additions": 21324, "deletions": 10581}}, {"node": {"oid": "e22ae49bcb8bf857d650042f0f9ebe06057c5c70", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/pavignol/1234", "author": {"name": "Patrice Vignola", "email": "pavignol@microsoft.com", "user": null}, "additions": 21324, "deletions": 10581}}, {"node": {"oid": "b4082e370bbcaaea156a0e2369ed7bb995d84423", "messageHeadline": "Merged PR 5557382: Revert \"Merged PR 5551793: Merge with latest ORT m\u2026", "author": {"name": "Patrice Vignola", "email": "pavignol@microsoft.com", "user": null}, "additions": 10603, "deletions": 21028}}, {"node": {"oid": "da952a9a20be79db5799680213ae3f9363034c7d", "messageHeadline": "A list of changes in transformers tool (#6224)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 29, "deletions": 9}}, {"node": {"oid": "ac5ca2bbe0fbfeecd50b1fdc0ca884831fd9d629", "messageHeadline": "fix data_ptr assertion error for past_sequence_length=0 in GPT-2 (#6284)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "7fc827a8a1c43e375ef0307576a75e61a194f761", "messageHeadline": "Fix Min/Max CPU kernels for float16 type (#6205)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 161, "deletions": 6}}, {"node": {"oid": "a72fcbd5fcca050be6c7689f09225109fd7b98c1", "messageHeadline": "Add helper to compare model with different precision (#6270)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 125, "deletions": 0}}, {"node": {"oid": "04287ec77025107e8ed9556f9d7a8aac29885a10", "messageHeadline": "Increase timeout for Linux GPU CUDA11 build. (#6280)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c10948699b28ad8149a32ab0b5a55551aae5da0e", "messageHeadline": "Rename MakeString and ParseString functions. (#6272)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 76, "deletions": 70}}, {"node": {"oid": "18ef0fafc40e9e1ca03baf853ea41391e039b81e", "messageHeadline": "Merged PR 5551793: Merge with latest ORT master", "author": {"name": "Patrice Vignola", "email": "pavignol@microsoft.com", "user": null}, "additions": 21028, "deletions": 10603}}, {"node": {"oid": "b80e8ce6a5df73423f5421918e64db384f3a830f", "messageHeadline": "rename past to past_key_values for GPT-2 (#6269)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "481a2cdf614458b550612176655656c4ec69b112", "messageHeadline": "Add script to preprocess python documentation before publishing (#6129)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 99, "deletions": 2}}, {"node": {"oid": "d761571afcbc50be94b2ffdfcaa09cfd9e2eaf5e", "messageHeadline": "Deprecate Python global configuration functions [Part 2] (#6171)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 937, "deletions": 370}}, {"node": {"oid": "a92e762f2260df9377babf9e3e97a294bd20c07b", "messageHeadline": "ci pipeline tests for ortmodule (#6268)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 251, "deletions": 1}}, {"node": {"oid": "bbc9ed908afc50a09952dca2e6af116541536a5c", "messageHeadline": "Fix VS 2017 build break (#6276)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "127afe3b097ab27d1fe9912d2ff7f2051315777e", "messageHeadline": "Device handling fixes in ORTModule (#6187)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 360, "deletions": 57}}, {"node": {"oid": "431604ef890ea5075a2a8245ad47e1a4608bf18b", "messageHeadline": "add bfloat16 to gathergrad type constrains (#6267)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2347de4a9e88391268e91b6abab84c908b09f82a", "messageHeadline": "Fix Linux/Mac error message on input type mismatch (#6256)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 104, "deletions": 28}}, {"node": {"oid": "d42399e1b07ce61e95aae88bc6b6ea5dcaae2011", "messageHeadline": "Allow querying a GraphProto's doc_string as part of ModelMetadata (#6\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 182, "deletions": 45}}, {"node": {"oid": "eea3806db1dbae7e0c46f6f9589b3b93982a9e10", "messageHeadline": "model parallel refinement (#6244)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 934, "deletions": 880}}, {"node": {"oid": "addb4b8c2b0331d0baa0599676200580a1c80637", "messageHeadline": "Liqun/speech model loop to scan (#6070)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 420, "deletions": 66}}, {"node": {"oid": "e0f2a12c2ca4224b251487568e855361a181a811", "messageHeadline": "ortmodule ci pipeline setup (#6251)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 127, "deletions": 0}}, {"node": {"oid": "ce6161cf672b6594a81bbdd9b99ec8c7f973829a", "messageHeadline": "Add MakeStringLite which uses current locale, update some MakeString \u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 61, "deletions": 47}}, {"node": {"oid": "493bf931c5862868abf431008ce2619cebf4740d", "messageHeadline": "Add the Concat Slice Elimination transform, fix constant_folding tran\u2026", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 422, "deletions": 0}}, {"node": {"oid": "6fd9d34bb0aa6241a6f4e5f030efe5927030dd12", "messageHeadline": "Remove a debug log in provider_test_utils.cc (#6200)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "93bf7c4d52bc96fb3ffbba39175b5ce7db3e01ae", "messageHeadline": "Documentation for distributed CI tests pipeline (#6140)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 56, "deletions": 1}}, {"node": {"oid": "c8de3f355a1859b65f95bc8a5ed1ee938f4a45c8", "messageHeadline": "Refactor EP Perf Tool  (#6202)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 169, "deletions": 136}}, {"node": {"oid": "46e0e4e69f53b5d70dac14c2d0609f7ee66ed603", "messageHeadline": "Tune BiasGeluGradDx kernel in approximation mode to avoid tanh(...) o\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 47, "deletions": 0}}, {"node": {"oid": "ffb4b628269ca28fc39993d7758ffe1827f9d19f", "messageHeadline": "Fix allocator issue for TensorRT IOBinding (#6240)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 25, "deletions": 8}}, {"node": {"oid": "1685167e46782ac4a61066f44d7a1ca8d875e9c2", "messageHeadline": "Update manylinux docker image to the latest (#6242)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 131, "deletions": 81}}, {"node": {"oid": "d5cb17c67903c50aa1bb370e9470c8c630531f14", "messageHeadline": "Update BUILD.md", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "cd14c1af29665c2941e6da78ea38b9f035dec94d", "messageHeadline": "Support double for operator ArgMin (#6222)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 58, "deletions": 8}}, {"node": {"oid": "84addcd2cf332ab898d4e2125523ec79a9eb0fb8", "messageHeadline": "Support double for operator ReduceMean, ReduceLogSumExp (#6217)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 265, "deletions": 0}}, {"node": {"oid": "5968a91ea66d7a1f0568b042b8393f0b4c9c8a55", "messageHeadline": "Support double for operator Gemm + fix bug in gemm implementation for\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 264, "deletions": 146}}, {"node": {"oid": "70e2f96ef43f60f121f230ec445c56b2d54e94e4", "messageHeadline": "Support double for operator TopK + fix one bug in TopK implementation\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 436, "deletions": 152}}, {"node": {"oid": "ecb2e119e468facdd04a49cf04e2d9d70084ad95", "messageHeadline": "MLAS: handle MlasGemm(M/N/K==0) cases (#6238)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 264, "deletions": 246}}, {"node": {"oid": "4cc2ffef2129cc4fe861d0af2899c42ceed079ce", "messageHeadline": "Support MLFloat16 type in Pow opset-12 CUDA kernel (#6233)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "39a988ce1c24c1ac5ab062544092507b924cffae", "messageHeadline": "Upgrade build.py to assert for python 3.6+", "author": {"name": "William Tambellini", "email": "wtambellini@sdl.com", "user": {"login": "WilliamTambellini"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c15a8587458fc747ddd60e3db0b357b8f3300003", "messageHeadline": "Update the readme file", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3911105f0916196e03190df76aeed91f34860d6f", "messageHeadline": "Remove python 3.5", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 16}}, {"node": {"oid": "1b23b2870602d9e41901b8ac0f01c6949c6b5130", "messageHeadline": "Remove MKLML/openblas/jemalloc build config (#6212)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 28, "deletions": 528}}, {"node": {"oid": "5c584b26363ab70741ce8693b5b4267ab059f72d", "messageHeadline": "Removed executor todo that looks dead. (#6234)", "author": {"name": "Michael Giba", "email": "michaelgiba@gmail.com", "user": {"login": "michaelgiba"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "bbb6b416f016bfa68c3cb78e5c1a323e8a25f8ba", "messageHeadline": "Fix ImportError in build.py (#6231)", "author": {"name": "Michael Goin", "email": "mgoin@vols.utk.edu", "user": {"login": "mgoin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "df7e2f3c1edee91f11267f3946ef5a30d2d7c6e8", "messageHeadline": "Support double for operators Relu, Tanh, Sigmoid (#6221)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 116, "deletions": 75}}, {"node": {"oid": "111ac299cc26d0b7dfca68e500855e927653c157", "messageHeadline": "Support double for operators Where, LpNormalisation (#6034)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 131, "deletions": 78}}, {"node": {"oid": "2d09db67b48c322ceb63cfffc4342b8586a7b2ec", "messageHeadline": "Support double for operators Log, Reciprocal, Sum (CPU) (#6032)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 161, "deletions": 18}}, {"node": {"oid": "8a0f5c50abba92175912fcda813512d8d7386499", "messageHeadline": "Minor change to improve performance for operator Pad. (#5537)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "7ccdfed1a680f44de5afac6950a1d96364f77362", "messageHeadline": "Remove most ROCm-specific element-wise code and reuse CUDA element-wi\u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 0, "deletions": 1418}}, {"node": {"oid": "52228a703cc5164fd571d0085af381bd8c3246ec", "messageHeadline": "Use TArray in AMD element-wise kernels, rather than manually copying \u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 172, "deletions": 202}}, {"node": {"oid": "1fc7f92f254f99092bc69e85b817469a281adad6", "messageHeadline": "Fix a memory leak in test_inference.cc (#6201)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "73479969427a33ad679607d7e4d14f3c142c22e2", "messageHeadline": "Openvino ep 2021.2 (#6196)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 1109, "deletions": 77}}, {"node": {"oid": "0494a0f95fae2bd381cc4ef2aaef5b49f18d9577", "messageHeadline": "Add ability to skip GPU tests based on GPU adapter name (#6198)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 101, "deletions": 3}}, {"node": {"oid": "c562952750f66b59329e1b088a4056f6be84a5f7", "messageHeadline": "Dockerfile to build onnxruntime with ROCm 4.0", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 195, "deletions": 0}}, {"node": {"oid": "21395f8e2465c6ab8ce7ed35880f7179ff537556", "messageHeadline": "Implement comparing outputs that are sequence of maps of strings to f\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 197, "deletions": 34}}, {"node": {"oid": "a8b482681a9ffd9595813c1e890a2b01a5db6d57", "messageHeadline": "Clean up checkpoint tests to use the new checkpoint functions (#6188)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 530, "deletions": 947}}, {"node": {"oid": "04b3e0ef5eb7916f9806b89e3e5e41172ccd3dbb", "messageHeadline": "Condition fix in Resize operator (#6193)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "fc27074baef898ece75a9444a8a7605ed508f2d1", "messageHeadline": "Implement ScatterND for CUDA EP (#6184)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 307, "deletions": 33}}, {"node": {"oid": "945fae8f56a11f29b5561837a017272a35a9566d", "messageHeadline": "Lochi/quantization tool for trt (#6103)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 1107, "deletions": 21}}, {"node": {"oid": "234e94b4e1aef7b11f9c939b5bf9f8b1e51d6d64", "messageHeadline": "Add Status.csv to EP Perf Tool (#6167)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 167, "deletions": 50}}, {"node": {"oid": "67ac6ae4e0a6e6d123f5d526e64d52bd56ef9ff0", "messageHeadline": "Tune fast Gelu to use exp(x) instead of tanh(x) on Rocm platform (#6174)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 21, "deletions": 2}}, {"node": {"oid": "53307a5f2ea5a665a80fc55b40965c95f70bd1f6", "messageHeadline": "improve perf for softmax (#6128)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 406, "deletions": 0}}, {"node": {"oid": "ea9cfa554a9de5abc9a5747c6f9ba327a482c7b6", "messageHeadline": "Add usage details of unified MCR container image (#6182)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "201d0dbb1a71998997ed1d59a3e78d873e7a1806", "messageHeadline": "Android coverage dashboard (#6163)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 85, "deletions": 15}}, {"node": {"oid": "f874260b9e0d26a603b64e15da5a454a4b67c229", "messageHeadline": "Backend APIs for checkpointing (#5803)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 570, "deletions": 129}}, {"node": {"oid": "2da8060f3407d5e90af470a004ff72732b8ad82d", "messageHeadline": "Helper for compiling EP to generate deterministic unique ids for use \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 225, "deletions": 21}}, {"node": {"oid": "cd3a5acca04e90118a897f8ea6bc52c38116c03e", "messageHeadline": "Update get_docker_image.py to enable use without image cache containe\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 68, "deletions": 42}}, {"node": {"oid": "11b0a5401e69bdfb92537a5b067f12c34d00cc15", "messageHeadline": "Fix typo in BERT pretraining script (#6175)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "bbb52e9274e79c83c0c89dbb50e2ddfeb7879cd2", "messageHeadline": "[NNAPI EP] Enable per-channel quantization for QlinearConv  (#6155)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 410, "deletions": 79}}, {"node": {"oid": "39aedbc97fc8e4d945792c81528e4950e323e919", "messageHeadline": "aggregate model states only for the case when mixed precision was tru\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 109, "deletions": 10}}, {"node": {"oid": "86493e6d0c1b9bbab1597683f51932c89db13dd1", "messageHeadline": "Update documentation for contributing a PR and add deprecation notice\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 13, "deletions": 9}}, {"node": {"oid": "824ef9a1de494682494dd9d896587eb30e310d94", "messageHeadline": "Don't try to bind unused inputs in the Training frontend (#6166)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 30, "deletions": 7}}, {"node": {"oid": "adc2071043dffe7fc0739b292afd4502483a51b5", "messageHeadline": "save_checkpoint, load_checkpoint and aggregate_checkpoints (#6136)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 717, "deletions": 56}}, {"node": {"oid": "c339bb2da9a07ac080863be938ccf51df70b61af", "messageHeadline": "Remove ignored build warnings for pybind on Mac (#6165)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 0, "deletions": 9}}, {"node": {"oid": "98d8a3e335e3f9d65b4f8aa54643d345b9493ad7", "messageHeadline": "Revert \"Fuse MatMulIntegerToFloat only when scales are scalar (#6008)\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 6, "deletions": 21}}, {"node": {"oid": "34725ae520c47a66ebef3c40af6a251ff952706d", "messageHeadline": "Bugfix for topk cuda kernel (#6164)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 46, "deletions": 8}}, {"node": {"oid": "dec703b62de498d995563af0b33f1647e26f30ab", "messageHeadline": "Update TensorRT-ExecutionProvider.md (#6161)", "author": {"name": "Jay Rodge", "email": "jayrodge@live.com", "user": {"login": "jayrodge"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "32c67c2944d86e62de5c15230b4ab311b8f932d6", "messageHeadline": "Deprecating Horovod and refactored Adasum computations (#5468)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 3178, "deletions": 1422}}, {"node": {"oid": "efa1b0d8645c1cc697252cc11c222805fbaee2fb", "messageHeadline": "Minor fix to satisfy c++14 (#6162)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "36c03b32e98645d64444124ec11783138b9403f4", "messageHeadline": "Using a map of of ops to stages as input of partition function. (#5940)", "author": {"name": "Juliana Franco", "email": "jufranc@microsoft.com", "user": null}, "additions": 1683, "deletions": 261}}, {"node": {"oid": "503b61d897074a494f5798069308ee67d8fb9ace", "messageHeadline": "MLAS: add NEON version of int8 depthwise convolution (#6152)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 103, "deletions": 16}}, {"node": {"oid": "0fa04bdc50a0dfbd3a3c99ced03aa824457516ae", "messageHeadline": "Fix clean_docker_image_cache.py detection of image pushes. (#6151)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 17, "deletions": 13}}, {"node": {"oid": "344a2a8ee5b3e8d626a030381b95ca701237c890", "messageHeadline": "Revert \"work around of the build break in mac (#6069)\" (#6150)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 8}}, {"node": {"oid": "72505622716feed356ea62c7a5d2a791ee151ffd", "messageHeadline": "Fix edge case in BFCArena where allocation failures could lead to an \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 22, "deletions": 1}}, {"node": {"oid": "82690486c17f4b7cd4f2a10595d2302ffbd47f7a", "messageHeadline": "Partition initial optimizer state for Zero-1 (#6093)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 335, "deletions": 45}}, {"node": {"oid": "8fd085801a39d299ff24962506fff82bc64fc001", "messageHeadline": "Add gradient registration for Abs. (#6139)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 13, "deletions": 0}}, {"node": {"oid": "aa49e476b09b7ea20bee612c97765a5b0773a16c", "messageHeadline": "Fix TensorRT kernel conflict issue for subgraphs of control flow oper\u2026", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 11, "deletions": 10}}, {"node": {"oid": "0978d2bfbee8ef20a5273577da6273a882dea5c0", "messageHeadline": "Fix CUDA test hang: (#6138)", "author": {"name": "Yateng Hong", "email": "toothache9010@gmail.com", "user": {"login": "toothache"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "b648bf641f78dda1a3ae189bd11de63ae26b14c2", "messageHeadline": "nnapi add min max support (#6117)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 167, "deletions": 13}}, {"node": {"oid": "b8c8fe91f55343925a5580b4cea204a544f1744d", "messageHeadline": "ort's to_dlpack.", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 332, "deletions": 27}}, {"node": {"oid": "939cc9b4106ce38c284009fc34127a1c6d722e09", "messageHeadline": "Enable running the mnist_training sample without cuda (#6085)", "author": {"name": "George Nash", "email": "george.nash@intel.com", "user": {"login": "georgen117"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "ac62cf80581d056f230d79bb59a3cd104232cd19", "messageHeadline": "Unify IExecutionProvider and IExecutionProviderFactory interfaces (#6\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 909, "deletions": 1008}}, {"node": {"oid": "980a93c164a9a3a0f724b9790b22692b9d896f95", "messageHeadline": "Model Fusion For Bart (#6105)", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 66, "deletions": 25}}, {"node": {"oid": "297c8248072dae8fb7f2e9e6f9b98c48697e310a", "messageHeadline": "remove dnnl_dll_path from post build copy (#6142)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "64709b1335380259b6b6e40cd02f5c3aeee7b38d", "messageHeadline": "Deprecate Python global configuration functions [Part 1] (#5923)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 778, "deletions": 487}}, {"node": {"oid": "c4f827bee1a112538458cbdb965673520ddb01c4", "messageHeadline": "remove initializers from original graph", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 26, "deletions": 8}}, {"node": {"oid": "f7f435fc27932d16e39d0763324886a97ceed1b2", "messageHeadline": "Improve dynamic axes to work without data descriptors", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 39, "deletions": 33}}, {"node": {"oid": "7729bb3c8d19956f4fc26ee0729dc1d8285b9eb1", "messageHeadline": "Add initial dynamic axes support", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 107, "deletions": 100}}, {"node": {"oid": "004632ff8d500e86044c037a3b98f90714aa8fa6", "messageHeadline": "TEMP: Add support to measure method execution time for perf improvement", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 35, "deletions": 4}}, {"node": {"oid": "e5fdb455edc7693d0177dd796ac5309eaf654e0f", "messageHeadline": "Improve performance by running ApplyTransformers on gradient graph", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "e986ae5f86acd7b7f2bde71fb89717e9e0d72cf6", "messageHeadline": "Remove dead code", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 0, "deletions": 28}}, {"node": {"oid": "41b88ce91dc9c750ecb3c21c8249d0a16fe0df61", "messageHeadline": "Remove initializers from forward ONNX graph", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 9, "deletions": 28}}, {"node": {"oid": "07f5ae95e5b6293c144b0c0556fcf3cfb79ae2c4", "messageHeadline": "Refactor MNIST and BERT classifier to add time measures", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 86, "deletions": 21}}, {"node": {"oid": "395e082bc30ef9095b7c07b1d26bb4fc752d9799", "messageHeadline": "Remove (unnecessary) gradient graph from frontend", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "4d9267e102b3ab6b055a3c87dc0ef238bb97fac5", "messageHeadline": "Add IO binding support, which allows CUDA training", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "f13c2a61d5efc3e5cf7cdf529b45cd574487a7c2", "messageHeadline": "Refactor IObinding", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 144, "deletions": 306}}, {"node": {"oid": "39ac95b2fc08ec4d75b6f9f59abe74de22c4907d", "messageHeadline": "add io binding", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 328, "deletions": 92}}, {"node": {"oid": "ff79e8743fbf71e7fa54ff618ef29383f19c5338", "messageHeadline": "Add support to BERT fine tuning (MVP 3)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 179, "deletions": 305}}, {"node": {"oid": "78831d009b6e3b6d87612e2ee0c597b7a4299649", "messageHeadline": "Add list of initializer gradients to the backend training graph spliter", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "60b6e2683ff45dabcf292d9d48601c2c19c1d8e2", "messageHeadline": "bugfix", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "f6a8d2aa5ff709e06670bbf4770c8e796bfc30fa", "messageHeadline": "split graphs info", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 232, "deletions": 245}}, {"node": {"oid": "cfd57c01361dbe3f85578f6d1873540af7a7637c", "messageHeadline": "fix input order, and input grad.", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 37, "deletions": 19}}, {"node": {"oid": "e759da178d039bdad9fc3825da53c78257e95446", "messageHeadline": "bugfix for graph inputs and outputs.", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 46, "deletions": 12}}, {"node": {"oid": "b7564d07328a1d67022f8c1ddc07a8208350e2d0", "messageHeadline": "Refactor after Vincent work on splitting on backend", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 18, "deletions": 141}}, {"node": {"oid": "6d8fde83246228b4d42c7be32a7efbe94590b24c", "messageHeadline": "sample code change.", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 10, "deletions": 13}}, {"node": {"oid": "934feb0c9931a4be421a6c66c52ab06cac310067", "messageHeadline": "gradient graph split in backend.", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 271, "deletions": 23}}, {"node": {"oid": "ea5871ac159b0357a21fadb9889a73fa29aab2cc", "messageHeadline": "Change DropouGrad.input[1].input_type and del logits_grad from backwa\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 18, "deletions": 12}}, {"node": {"oid": "f1dc6e4007f4f6662292da3c041594f05f184273", "messageHeadline": "Refactor BERT classifier fine tune for better debugging", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 210, "deletions": 211}}, {"node": {"oid": "d4917f2d6599c12a52bd76b89fac8c8c7cb3c3cf", "messageHeadline": "Hard-code input types for DropoutGrad on BERT", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 54, "deletions": 22}}, {"node": {"oid": "3b267d1d6032a13d71cc58a4897ae0f68556838f", "messageHeadline": "Add BERT classifier example", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 379, "deletions": 1}}, {"node": {"oid": "30042b6e0ec63f9fa726c778e68886a8144632de", "messageHeadline": "Update InferenceSession usage to match master", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "8b0ade0e834b7cc70d3784a6e562bc5d2d637469", "messageHeadline": "Integrate automatic graph split into ORTModule", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 501, "deletions": 284}}, {"node": {"oid": "c36c8e14a7d6172735eaa144f7c7cfa2c31cb6b1", "messageHeadline": "refactor", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 149, "deletions": 65}}, {"node": {"oid": "26e6d6d00475dee4200cecc0ffcd77707c30a306", "messageHeadline": "module transformer", "author": {"name": "Vincent Wang", "email": "weicwang@microsoft.com", "user": {"login": "er3x3"}}, "additions": 199, "deletions": 0}}, {"node": {"oid": "3524fb04e80b83cc2cd3d3ee999405c813088be2", "messageHeadline": "Add working example for MNIST (MVP)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 183, "deletions": 159}}, {"node": {"oid": "f1b5c25b2d96a3417b8bc227e8198f3d40abed55", "messageHeadline": "Improve example to display grads before and after optim step", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 47, "deletions": 44}}, {"node": {"oid": "f06cafdebd88bfb9d144bda0f87b722f45777815", "messageHeadline": "Fix path on test script", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "56ca4ab05b9719bc86bb2bbe4575bbf08c270922", "messageHeadline": "Add flag to allow pytorch-only or ORT flexible api runs", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 111, "deletions": 81}}, {"node": {"oid": "d4449d86b95345465d5b420487b0e17645652e94", "messageHeadline": "Add script to run Flexible API MVP PoC", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 97, "deletions": 31}}, {"node": {"oid": "e71e08851a84c4c6417c04f79710ebb6e6b7bee6", "messageHeadline": "Basic plumbing for backward pass. Not fully working", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 318, "deletions": 134}}, {"node": {"oid": "77cefcd6c2435fd39be7cf526324ea206de21f21", "messageHeadline": "Perform forward pass using training graph with intermediate outputs", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 225, "deletions": 14}}, {"node": {"oid": "11b69f141eaca39ba5fb56eec77e6822e17cf9cc", "messageHeadline": "Forward pass using InferenceSession on exported ONNX", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 275, "deletions": 0}}, {"node": {"oid": "a8d549e181534e80a519cced7ee3aedebeef1b1d", "messageHeadline": "Minor changes to AMD element-wise kernels to converge with CUDA eleme\u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 18, "deletions": 52}}, {"node": {"oid": "a9548283d0cf85b434897cbd62e9fd0e801bacdc", "messageHeadline": "Don't mark issues that are marked as enhancement as stale (#6134)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "9810b9e02b5d15e8c3edbe1ade44704e666127b4", "messageHeadline": "Reduce amount of compiled CUDA device code (#6118)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 610, "deletions": 580}}, {"node": {"oid": "a6a23db1303ee77582aff85d13218e4c3e7e64fb", "messageHeadline": "Enable C# .NET5 for WinML  (#6120)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 64, "deletions": 486}}, {"node": {"oid": "eb5c1f0fcc6038f2743e7f5d2c1c43a014bdfdcb", "messageHeadline": "Unify activation and initializer alignment value (#6109)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 64, "deletions": 40}}, {"node": {"oid": "cde723a1361ab2c66879cebb3515d5e5eef763fa", "messageHeadline": "Liqun/move nightly pl to linux multi gpu v100 (#6024)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 49, "deletions": 17}}, {"node": {"oid": "dd2e5a1a052c0ef8241af6fe38404dc2b4d1310d", "messageHeadline": "state_dict and load_state_dict for ORTTrainer (#6095)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 806, "deletions": 4}}, {"node": {"oid": "d4dddd99d9e42d35247bbcbbb0c87feffb47f633", "messageHeadline": "Bump ini from 1.3.5 to 1.3.8 in /nodejs", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "c755ca0b71412b196086af840132e8f6a7082254", "messageHeadline": "Honor auto_pad attribute in ConvTranspose  (#4271)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 233, "deletions": 96}}, {"node": {"oid": "6cb5d3ac09ec579791c58a13063f90e616ddaeb1", "messageHeadline": "Fix multi-tensor LAMB reduction to be deterministic (#6028)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 236, "deletions": 104}}, {"node": {"oid": "c8ac34d6a5d5fc55871da735171f6d93f80a714c", "messageHeadline": "Fix DEBUG_NODE_INPUTS_OUTPUTS test by putting it in a separate proces\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 87, "deletions": 176}}, {"node": {"oid": "a53f4dd3799f9e5741184d2051c5347c578bb522", "messageHeadline": "Introduce VariadicAlias, remove hardcoded alias limits (#6106)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 54, "deletions": 54}}, {"node": {"oid": "38c49c2483706a5bc08885c58d658fb8172b9f4b", "messageHeadline": "Make ROCM and CUDA reduction_all code more similar.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 24, "deletions": 43}}, {"node": {"oid": "1eb146f561fab04445f36ee7f18b967433ecccdb", "messageHeadline": "Implement conversion from ORT String to WinML Tensor String (#6097)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 55, "deletions": 3}}, {"node": {"oid": "3b44eee2414b45c0f56b6c2874e15a3d4d4ea227", "messageHeadline": "Merged PR 5490057: Preserve history of merge", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "a12faed0a06efcc7b01f31cac50ea7333b496dad", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into 054fb4d3-revert-f\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 3272, "deletions": 1711}}, {"node": {"oid": "c4429952aaa5265b51e1516e6aeaee674e5578e2", "messageHeadline": "Revert \"Merged PR 5490021: Merge latest github master into dmldev branch", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 1711, "deletions": 3272}}, {"node": {"oid": "054fb4d3f69d8c11945bc55e04f6afd2271aee2f", "messageHeadline": "Merged PR 5490021: Merge latest github master into dmldev branch", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 3272, "deletions": 1711}}, {"node": {"oid": "8bcb5fd119fcd29d16a36979eecfcaaeefb04e60", "messageHeadline": "Add skip test reason for onnx model zoo models and tier 2 models (#6081)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 13, "deletions": 19}}, {"node": {"oid": "753af576c402a9069decebe24d1b74adbc008292", "messageHeadline": "If building inbox, hook up winrt_activation_handler for WinML Tests (\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 45, "deletions": 1}}, {"node": {"oid": "e945b5fcf649f2301e72d64fcd0df9d3c994fade", "messageHeadline": "adding fp16 support for topk cuda kernel (#6082)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 53, "deletions": 16}}, {"node": {"oid": "7ddeafdfcc498681f3a665ffa609646bf52edeaf", "messageHeadline": "Add ReduceL2Grad and ClipGrad (#5970)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 128, "deletions": 2}}, {"node": {"oid": "404982ded525c41d8e379112874732fc534e946a", "messageHeadline": "Enable varied input type for custom op (#6066)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 179, "deletions": 32}}, {"node": {"oid": "cc47cfcb314edeef761b48cb19feb7515f095f42", "messageHeadline": "Update AMD transpose to match CUDA transpose.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 81, "deletions": 393}}, {"node": {"oid": "abdbb5fc8441f56c69d6c4ed5a3f448525158f41", "messageHeadline": "Reduction kernel optimization (#6088)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 24, "deletions": 26}}, {"node": {"oid": "9e26e59a3744c3c47e442d38d0568fbedd290bef", "messageHeadline": "Deprecate opsets <12 for training. (#6027)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d95fc5e84956967f0f5bb07c8d1321d1caf43ed0", "messageHeadline": "clean un-used code. (#6059)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 1, "deletions": 299}}, {"node": {"oid": "27051157323dbe5c95b144b4afcdef7b5c536d40", "messageHeadline": "add dockerfile for ROCm3.10 and update BUILD.md for ROCm EP (#5821)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 278, "deletions": 12}}, {"node": {"oid": "b1a75d0e98e947497725e79f8f939c9b08c599a2", "messageHeadline": "Enable passing initial optimizer state while creating training sessio\u2026", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 973, "deletions": 293}}, {"node": {"oid": "7a43fa00285ce8a923bf1a1d7907818cdaa0d063", "messageHeadline": "Fix AllReduce kernel for contiguous buffer (#6064)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 12, "deletions": 5}}, {"node": {"oid": "e3574867077979f7e50cc54175786baa7bde2c47", "messageHeadline": "Fix build definition template typo, add logging (#6065)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "523d18719348072b6e31e0e9f00780200cc47440", "messageHeadline": "save data to and load data from an hdf5 file for checkpointing (#5975)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 313, "deletions": 0}}, {"node": {"oid": "3e81711a1396bc4eb1b0a5775036a550d93c59da", "messageHeadline": "Update version to 1.6.0 (#6041)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 24, "deletions": 12}}, {"node": {"oid": "f68a256140402b434939e91fb9b1cddec1389250", "messageHeadline": "Android code coverage (#6061)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 118, "deletions": 6}}, {"node": {"oid": "e35211c0ff2fc898a753f315c74a2051ad672a8c", "messageHeadline": "Fix AMD GPU pipeline by adjusting reference /opt/rocm-3.9.0 => /opt/r\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "2c5ba9ab00b63fbea46c71e0ecaabe83cb0b4ca1", "messageHeadline": "Bump up API version for 1.6 release (#6076)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "3cae28699bed5de1fcaadb219fa69bae0fc3cee8", "messageHeadline": "work around of the build break in mac (#6069)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "fa06be21330304c38c8627673348840eef2449a1", "messageHeadline": "Support export >2G model when using optimizer.py only (#6014)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 11, "deletions": 2}}, {"node": {"oid": "b348538c8ab992937e1178e1895a3c2d1fe4707a", "messageHeadline": "Update build docker image cache cleanup (#6048)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 45, "deletions": 17}}, {"node": {"oid": "51fbe87b9bd66e95db2d1a324997cbc15db9ee46", "messageHeadline": "Update profiler tool to support gpt2 and longformer models (#6011)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 370, "deletions": 188}}, {"node": {"oid": "925879a8b0db55e7b9d88322305ff57002d0f026", "messageHeadline": "Remove python 3.8 Windows GPU build from python packaging pipeline (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 6}}, {"node": {"oid": "020efc90028af5a16607ec35ab266002ca83e600", "messageHeadline": "fix windows cuda support for python 3.8 +  (#6046)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 11, "deletions": 14}}, {"node": {"oid": "7cebf76a46b4d44de0a11c13b815c6d929b67ae4", "messageHeadline": "Improve checkpointing for Zero stage 1 (#5478)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 571, "deletions": 361}}, {"node": {"oid": "0afbdfd81ca81509dea957bc7335c4fdc150f27b", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into HEAD", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1755, "deletions": 416}}, {"node": {"oid": "a046ef133aa18bda7b7ec9eeedfec4800f452d45", "messageHeadline": "Update api_summary.rst (#6038)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "d5e8c48e546cf19516622a7df41268f801eb10c8", "messageHeadline": "Bump highlight.js from 10.2.1 to 10.4.1 in /nodejs", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d8139814fd03a6e297c778489036e66c3df94887", "messageHeadline": "Clean up builds (#6015)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 41, "deletions": 55}}, {"node": {"oid": "00f43a3a689c6ca39651a24eded96f0a7687e728", "messageHeadline": "add missing iclosable interface (#6036)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "14f6eb14b1a0da7fc580f550f660ef248e8a552f", "messageHeadline": "Use __launch_bounds__ workaround, rather than limiting threads to 256\u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 1, "deletions": 149}}, {"node": {"oid": "98ea7372d36175b4e976a4fa6e65e314a5594ddb", "messageHeadline": "Re-enable Lamb unit tests for AMD", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 0, "deletions": 13}}, {"node": {"oid": "245d43615d592e84cb6a6b6802dd383952b2581b", "messageHeadline": "Fix AMD multi-tensor implementation.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 151, "deletions": 9}}, {"node": {"oid": "6572a4d306ac20d7874af985a9db062904198b42", "messageHeadline": "Disable Python 3.9 for training Python packaging build. (#6012)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "cdb91208a311a6b842725c0dbe4f6d9a106eaf3e", "messageHeadline": "longformer onnx conversion and benchmark tools (#6007)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 565, "deletions": 0}}, {"node": {"oid": "3b198c96146cc4064ac5135d0cdd9a712735c0a7", "messageHeadline": "Support Fusion for 1 and 2 Inputs Bert Models Converted From tf (#5993)", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 122, "deletions": 43}}, {"node": {"oid": "c86a1e5c13057b29c72c839abeb6f97e9751c041", "messageHeadline": "Fix Flaky orttraining tests (#5977)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 17, "deletions": 4}}, {"node": {"oid": "2878e8eb2e27a5f603f029e4922eef30028a5a4e", "messageHeadline": "Fix nuget build error (#6009)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 6, "deletions": 6}}, {"node": {"oid": "2b35f7d4f6b1d39170859fd772e710f894cb1765", "messageHeadline": "Fix build.py bug which prevents running some unit tests (#5990)", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 15, "deletions": 7}}, {"node": {"oid": "0acc3837eeb6137825f8033e03653f8a1bf11407", "messageHeadline": "Make operator TreeEnsemble 5x faster for batches of size 100.000 (#5965)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 301, "deletions": 91}}, {"node": {"oid": "524b9fa89997c2b45066a971861b465943c5ad67", "messageHeadline": "Initialize a structure in operator ReduceSum (#6005)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 15, "deletions": 5}}, {"node": {"oid": "648c9c7789767d4090e2a048b811cdf8e2158ddf", "messageHeadline": "Fix bugs for 1: Calibrator should check model inputs; 2: (#6017)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "bdd06f63109ddf85ead6f836f36f85bf4fb1bc9f", "messageHeadline": "Fix PR #5550 reverted in #5911 (performance improvment for operator T\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 322, "deletions": 102}}, {"node": {"oid": "f2dcba7afe0d42ebdaaef0c6cdf913a1156c9e98", "messageHeadline": "Fuse MatMulIntegerToFloat only when scales are scalar (#6008)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 21, "deletions": 6}}, {"node": {"oid": "4fdfbfd4b4628972d08ca861feaa7d9cbd635730", "messageHeadline": "Add int32_t support for DeQuantizeLinear (#5994)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 48, "deletions": 4}}, {"node": {"oid": "c727a28735c845b898f86011530044d5a8b6f986", "messageHeadline": "include gemm_helper.h (#5988)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b4e6cc59c713f9df0a6e877edd2128556c6e23c5", "messageHeadline": "skip the check for A channel (#5989)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "cdacee6696262d0c5431b6860bb0966afdc6eb0d", "messageHeadline": "[NNAPI] Support non-1d tensor for C of Gemm op (#5982)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 143, "deletions": 56}}, {"node": {"oid": "6846c665ff12d36290210436fd944646ec34d1d6", "messageHeadline": "Use loose version in build.py (#5998)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "897310f6fb709713298efea7ea772687de7fb67a", "messageHeadline": "Add suspend handler with new telemetry event for UWP scenarios (#5907)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 121, "deletions": 7}}, {"node": {"oid": "934b7882f093fa55c24d69449fd86edc14ca608e", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into dmldev_temp", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 12650, "deletions": 7027}}, {"node": {"oid": "6d642a3dbad00ca3e49db391b329c5514a89ed3f", "messageHeadline": "Replace direct pulls from image cache container registry with get_doc\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 394, "deletions": 463}}, {"node": {"oid": "c63e8cf7d71c9423bb5f7f00f69fb13b087f4a84", "messageHeadline": "Remove chronological starttime assertion in InferenceTest.cs because \u2026", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 3, "deletions": 9}}, {"node": {"oid": "396074d2a8ddcde5627f0eae8b63f51299c43b4a", "messageHeadline": "Fixing OrtEnvironment.getEnvironment() so it doesn't print a warning \u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 18, "deletions": 5}}, {"node": {"oid": "30c7fffbab8f8b38d9731f050c42363d5a2714bd", "messageHeadline": "Expand the documentation on using compiling EPs with a minimal build \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 149, "deletions": 55}}, {"node": {"oid": "46277bfd09bfa05f557a654a34697d7a1e99a350", "messageHeadline": "Remove survey link from README (#5979)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "5f516899bfe55cc132aa6ce2cb02c1f1a35cbc61", "messageHeadline": "optimize a bert model converted using tf2onnx (#5492)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 190, "deletions": 27}}, {"node": {"oid": "3323fb608278108b7ebd98c74b5716779e84056e", "messageHeadline": "Update docker files to put 'unattended-upgrades' in a right place(#5983)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 6}}, {"node": {"oid": "2ec211ea7b28bc7d1831c57ec345fecbea917609", "messageHeadline": "Support the cross compiling for Apple Silicon (#5974)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 31, "deletions": 13}}, {"node": {"oid": "fb310fba0c837310af40d87fc47c78586e362124", "messageHeadline": "Avoid adding non-existent inputs to new Event nodes (#5915)", "author": {"name": "Alberto Magni", "email": "49027342+alberto-magni@users.noreply.github.com", "user": {"login": "alberto-magni"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "2d9dcc4576f86b3eb6705dba1f8699eb52140b4e", "messageHeadline": "Add python 3.9 support (#5874)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 223, "deletions": 82}}, {"node": {"oid": "1852ade75dc57ad37834a92f57277819c2b142e3", "messageHeadline": "Enable the xcode build for Apple Silicon (arm64 MacOS) (#5924)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 24, "deletions": 15}}, {"node": {"oid": "45966d878ad425750f213cb5808e6bb49ea8b4f4", "messageHeadline": "Code review feedback", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "86e30a2db6f02ab4d2b92800fbf398f7ea641978", "messageHeadline": "Update CUDA IsAllFinite kernel", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 13, "deletions": 20}}, {"node": {"oid": "bd96f60888dc1eeb919fd18b3821d84c31e7b545", "messageHeadline": "Use CUDA's IsAllFinite kernel for ROCm", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 1, "deletions": 181}}, {"node": {"oid": "06ad516a5d38c479a62961b85fa9ae3ba4b3e9ab", "messageHeadline": "w (#5947)", "author": {"name": "Moshe David", "email": "mosdav165@gmail.com", "user": {"login": "mosdav"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5f5d4a10bdfb868beadc43c1c023fcf4ab7d0838", "messageHeadline": "[OpenVINO]Fix memory leak upon exception throwing (#5954)", "author": {"name": "Moshe David", "email": "mosdav165@gmail.com", "user": {"login": "mosdav"}}, "additions": 4, "deletions": 7}}, {"node": {"oid": "015fbb3dbb8547294171990e1104cdf101be4a7d", "messageHeadline": "Add support for Python 3.8+ on Windows when CUDA is enabled (#5956)", "author": {"name": "Ivan Stojiljkovic", "email": "17503404+ivanst0@users.noreply.github.com", "user": {"login": "ivanst0"}}, "additions": 16, "deletions": 2}}, {"node": {"oid": "e207589631a41d618f3e87c66de6b2f7e49ef52d", "messageHeadline": "[OpenVINO]Fix memory leak in `IsDebugEnabled()` under Windows (#5948)", "author": {"name": "Moshe David", "email": "mosdav165@gmail.com", "user": {"login": "mosdav"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "4afdced7755c5518b4345e2887e456ee683d5fe9", "messageHeadline": "[NNAPI EP] Update squeeze ops (#5946)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 78, "deletions": 51}}, {"node": {"oid": "d52b9aca680829715b69e2f8cd544a34d91bb526", "messageHeadline": "Enable scalar input/output for NNAPI EP (#5922)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 61, "deletions": 31}}, {"node": {"oid": "5fdd9f0fd26c8038aef8977af9ad554a5b8aa978", "messageHeadline": "Fix Python Linux GPU package name (#5943)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 7}}, {"node": {"oid": "27513d1fd796dbb623e50a9caed243ac9ae600ef", "messageHeadline": "Update BUILD.md with shared provider information (#5944)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 25, "deletions": 0}}, {"node": {"oid": "c4b55d29fe54605b582f373e9997642ff78f1aba", "messageHeadline": "Fix publishing pipelines. (#5942)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "1dbabb23626368dddac2a79e3df4852f8ea52cdb", "messageHeadline": "Update dockerfiles (#5929)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 26, "deletions": 42}}, {"node": {"oid": "c5b4d9091c1b454dffff502fbcd42928c6fdb547", "messageHeadline": "Fix a tiny issue in onnxruntime_unittests.cmake (#5901)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7546d251e06e753ab2a3c935a7aafe17ca682643", "messageHeadline": "Expose parameters in clean build Docker image cache build. (#5941)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 19, "deletions": 3}}, {"node": {"oid": "31a6be3d675d09fbe053e03d668695be1bd5fd89", "messageHeadline": "Add Longformer Attention Cuda Op(#5932)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1552, "deletions": 0}}, {"node": {"oid": "e39e82b43a3f1b632f8d54d5e9873c6d123826b9", "messageHeadline": "Bug fix for MaskRCNN and FasterRCNN (#5935)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 16, "deletions": 14}}, {"node": {"oid": "69b9368c9359b8941d3ffca1e0db788b5b0a38ca", "messageHeadline": "Add unit tests to identify configuration migration scenarios for chec\u2026", "author": {"name": "baijumeswani", "email": "bmeswani@microsoft.com", "user": {"login": "baijumeswani"}}, "additions": 1895, "deletions": 1}}, {"node": {"oid": "8168c919782a080490482f37100acdee94c2f9d2", "messageHeadline": "Sahar/fix documentation shared lib (#5926)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 12, "deletions": 21}}, {"node": {"oid": "87368655e23c61863d39d0eddb4a8d24612c0ba1", "messageHeadline": "Make NNAPI EP reject nodes with no-shape inputs (#5927)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 75, "deletions": 18}}, {"node": {"oid": "fddbd8935c853be532efe32155645561c195839a", "messageHeadline": "Adding Java support for getAvailableProviders and other small methods\u2026", "author": {"name": "Adam Pocock", "email": "craigacp@gmail.com", "user": {"login": "Craigacp"}}, "additions": 293, "deletions": 4}}, {"node": {"oid": "40926867c31c4a0118dd0a60b9e2dadefee6f237", "messageHeadline": "Add OpenVINO EP shared lib to Py Wheel (#5920)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "4d603e83d7b42087301dbb5fb042fe6972cca070", "messageHeadline": "Remove attention_past.cu and attention_transpose.cu from hipify to fi\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 2, "deletions": 0}}, {"node": {"oid": "8b83c51a35ea1673fb0af060ae59b3330cfd5628", "messageHeadline": "[Java] Initial Apple Silicon support (#5891)", "author": {"name": "Adam Pocock", "email": "craigacp@gmail.com", "user": {"login": "Craigacp"}}, "additions": 35, "deletions": 6}}, {"node": {"oid": "ee908eb0aa4ee6a94fcbae45ad6675738222d9fa", "messageHeadline": "Symbolic shape inference: fix rank for ConstantOfShape (#5912)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 2, "deletions": 1}}, {"node": {"oid": "c2d610066a460bc62bb030e4b6907f82c6617caa", "messageHeadline": "C#: Add CreateFromMemory to FixedBufferOnnxValue to allow bind user b\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 173, "deletions": 34}}, {"node": {"oid": "705d09316705f7431de7c6ed8151f8f6bbfe388c", "messageHeadline": "Update onnx (#5720)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "9b8189dd0a167667cadcc7c95bd2412ba9d71774", "messageHeadline": "Rework AMD CI pipeline to use pool AMD-GPU and disable more tests in \u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 62, "deletions": 129}}, {"node": {"oid": "846c5fb91753dd1020f4adb44f67db02c479924d", "messageHeadline": "Report arm64 minimal baseline binary size only for continuous integra\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 17, "deletions": 6}}, {"node": {"oid": "c49d5f1d982d3545c9a4be491322a12798530fb2", "messageHeadline": "Reenable skip flatten/reshape if it's Gemm's input (#5904)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 53, "deletions": 56}}, {"node": {"oid": "782303324ecf8b04f10e8ca9fe55af1fbf2a129b", "messageHeadline": "Revert \"Improves performance  of operator Transpose (#5550)\" (#5911)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 92, "deletions": 163}}, {"node": {"oid": "9992f0f812eb2d4d3ef9cdfc709ce820ff7015a9", "messageHeadline": "Implement QLinear GlobalAveragePool with sse2/neon. (#5838)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1368, "deletions": 30}}, {"node": {"oid": "d0a007e4f17086a26a06a3e7f9bff081b8b3269b", "messageHeadline": "Add info level logging for NNAPI EP partitions and number of supporte\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 11, "deletions": 4}}, {"node": {"oid": "4137c18d9b7013a77b71e34a0e4b6fcbdcf92990", "messageHeadline": "Add ORT minimal with NNAPI EP to Android CI (#5890)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 68, "deletions": 0}}, {"node": {"oid": "916410151c3bfa0e46a78b1d707106d537115786", "messageHeadline": "Fix for hetero multi python binding with new shared library (#5895)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "58ea7b35720772d021c0d8d8a8c7e7d9ac8b126c", "messageHeadline": "temporarily disable test (#5868)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "5e8fcda24a9463a988ccdb94ae081f62c0959c1b", "messageHeadline": "Build docker image cache fixes. (#5902)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 25, "deletions": 26}}, {"node": {"oid": "3d5b48a8948fb404c73c919a8305c8adcb5e813d", "messageHeadline": "remove use_cdn when loading pretrained model (#5900)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "208f4c1d3ca0c45b36d803b05e0b7bb1dbc9c995", "messageHeadline": "Azure ci pipeline for distributed environment tests (#5881)", "author": {"name": "baijumeswani", "email": "baijumeswani@gmail.com", "user": {"login": "baijumeswani"}}, "additions": 100, "deletions": 5}}, {"node": {"oid": "353e071b7e8cbefdb5bb6d5ee1a3a3fe976f147d", "messageHeadline": "Fuzz testing misc (#5862)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 10, "deletions": 71}}, {"node": {"oid": "f473dd295d241787b014423a9cbaf83fbe61a5fc", "messageHeadline": "Add QLinearMatMul(u8s8) (#5899)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 44, "deletions": 27}}, {"node": {"oid": "57c92066c2cbd384deecbe74e0950ab3e726718a", "messageHeadline": "Implement missing pieces for ARM QLinearConv support (#5894)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 645, "deletions": 315}}, {"node": {"oid": "d46dbeafd3bffe4ea153aedca3f81e3ef0538aa1", "messageHeadline": "Expose knobs to create and share (CPU) allocators across sessions in \u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 425, "deletions": 92}}, {"node": {"oid": "26e6ced17226eacd257f0efded60afd80a942d4b", "messageHeadline": "Temporary fix for Android CI failure (#5889)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 12, "deletions": 7}}, {"node": {"oid": "910bbfe1effa1031e8ee309353ee2f4e00caf7d0", "messageHeadline": "Support 3D attention mask (#5887)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 886, "deletions": 596}}, {"node": {"oid": "cc6e8fb7cce6f3699639e2d89533897d2b23967f", "messageHeadline": "Filter initializers for GraphViewer with IndexedSubGraph (#5884)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 39, "deletions": 4}}, {"node": {"oid": "ba739a80005dfd7eb1dbaac4d589d29c92393cfc", "messageHeadline": "Convert OpenVINO into a shared provider (#5778)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1222, "deletions": 1060}}, {"node": {"oid": "8c04ada0d0562f9f51e52c0b1fedffb61300601d", "messageHeadline": "Bump training docker image version. (#5428)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3738ca7e10e09460263207558138afd1e6b307c6", "messageHeadline": "Improve perf testing (#5760)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 107, "deletions": 48}}, {"node": {"oid": "f0142da59c3deff9d61befcb0d0ad32f41fa0436", "messageHeadline": "Add NNAPI to providers that can be used via the python bindings. (#5867)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 86, "deletions": 10}}, {"node": {"oid": "3970eb2e5da515c0884c430dc10ec3a522eb9014", "messageHeadline": "Add documentation on enabling/using NNAPI in a minimal build (#5879)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 54, "deletions": 6}}, {"node": {"oid": "c0471240b421a5f418d3ef1f71d3cb4e3d3ffde4", "messageHeadline": "Remove redundant lines of code (#5883)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "a622533ecca4a8415ba33cd8a628081f0248c022", "messageHeadline": "Support profile_file_prefix in python binding (#5864)", "author": {"name": "Takeshi Watanabe", "email": "take-cheeze@users.noreply.github.com", "user": {"login": "take-cheeze"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "5a0fdd353715ffdededd32640997d07152eec3aa", "messageHeadline": "Update custom_ops.cc (#5507)", "author": {"name": "KnightYao", "email": "KnightYao@users.noreply.github.com", "user": {"login": "KnightYao"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ceedf5630bd3fb9656e6c9e30ef3d2249dab565b", "messageHeadline": "Document all C# API pubic interfaces (#5853)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 743, "deletions": 101}}, {"node": {"oid": "ab9d4b366b7d6add87dd5ed18db299a5145884d3", "messageHeadline": "revert 262e9ef21dc319a8e7bece1f0d155364b17b019c (#5882)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 36, "deletions": 112}}, {"node": {"oid": "bef06dac93073c4a5434a9aa600e34b47e637127", "messageHeadline": "Automatically clean up build docker image cache. (#5843)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 253, "deletions": 10}}, {"node": {"oid": "6808bfefff03c2390d59c77eef9d196833821e7e", "messageHeadline": "Extend MatMulInteger fusion (#5871)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 273, "deletions": 194}}, {"node": {"oid": "44313970d33a3dcc1ebd3942167cfb66160f8cfc", "messageHeadline": "Enable scalar initializer support in NNAPI (#5875)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 81, "deletions": 52}}, {"node": {"oid": "00412a76e91d85d0ccba57ac8198bb4b73beab2d", "messageHeadline": "Exclude some training specific code from the minimal build. Cleanup s\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 161, "deletions": 125}}, {"node": {"oid": "b057b3d36e69118664152eca8eaedfb1b1dd72bc", "messageHeadline": "Enabled fp16 for input types (#5878)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 29, "deletions": 5}}, {"node": {"oid": "1068f3eb87693fa88d83e1d6193999c3a5000c29", "messageHeadline": "Use flatbuffers for INT8 calibration table (de)serialization in Tenso\u2026", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 160, "deletions": 6}}, {"node": {"oid": "fd6e7d9c5cadd017d37fe113f1bd934ec033018d", "messageHeadline": "Fix the arm64 build issue on some special OS for OpenVino (#5870)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4b17fd560f8c41a99de43052998a0c53136c834b", "messageHeadline": "Merged PR 5425356: DmlDev RI from GitHub master 2020-11-19", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1026, "deletions": 219}}, {"node": {"oid": "dfea92925c2cdb56e36d3f1d8a4f890e0e4b4d48", "messageHeadline": "Add calibration based INT8 quantization to TensorRT EP (#5842)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 296, "deletions": 61}}, {"node": {"oid": "ff58f621faf85897828a2509c4d2725522737a48", "messageHeadline": "Remove nGraph Execution Provider (#5858)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 89, "deletions": 2727}}, {"node": {"oid": "8f0bfe870bd910e490341f7d40fd543b92ac9561", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/dwayner/DmlD\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1018, "deletions": 181}}, {"node": {"oid": "261462be0dbc1fadeeeaaa1d612ca78987b133fa", "messageHeadline": "Change NNAPI runtime options to use uint32_t (#5863)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 35, "deletions": 35}}, {"node": {"oid": "4bb41d610bf2f1d822bb438631655931f83afc6b", "messageHeadline": "Address scalar expand to scalar case (#5854)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 26, "deletions": 0}}, {"node": {"oid": "e485805b8ded81219ce9309aa18d161d3e3a0717", "messageHeadline": "[NNAPI EP] Add infrastructure for Op Reduction (#5841)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 444, "deletions": 129}}, {"node": {"oid": "62508ef0e422791b3913515cf28305cd0d030551", "messageHeadline": "Revert \"Remove MKLML build config (#5559)\"  (#5855)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 513, "deletions": 17}}, {"node": {"oid": "cd07783ddec5ad8e0a0cef034e259c4da72b8af8", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 8, "deletions": 38}}, {"node": {"oid": "26db396b4bd8691f4bc2377fb5a379f6e4ea13be", "messageHeadline": "Reduce the number of CI build variants (#5856)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 38}}, {"node": {"oid": "62d4f75f186f44aef71d966b23a7a14d7b934011", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1595, "deletions": 965}}, {"node": {"oid": "47185b951349fe49120f9be3b13d708490efc78f", "messageHeadline": "reducealll2 cpu kernel (#5833)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 77, "deletions": 2}}, {"node": {"oid": "7196d4206fbc736330af819403146fdd0cad0e3d", "messageHeadline": "Adding Transpose3d and Transpose4d special case kernels for Rocm (#5837)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 282, "deletions": 7}}, {"node": {"oid": "b495ae81031e6d98199d15652ed1ff6bdcd56b8b", "messageHeadline": "ORT fuzz testing (#5771)", "author": {"name": "satyajandhyala", "email": "satya.k.jandhyala@gmail.com", "user": {"login": "satyajandhyala"}}, "additions": 618, "deletions": 485}}, {"node": {"oid": "84c1340f9bef299ca47fc389dabd6d821d4d3a10", "messageHeadline": "Refactor implementation of Tensor<T> and underlying buffer stores to \u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 548, "deletions": 401}}, {"node": {"oid": "63ee7b47712c8888fa6b85b05a3dcb3fea694016", "messageHeadline": "Merged PR 5420139: DmlDev RI from GitHub master 2020-11-18 #2", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 3008, "deletions": 1946}}, {"node": {"oid": "85f945a8757e47f6465d6167b8e629f8ab08e17a", "messageHeadline": "Regenerate CI build docker images (#5850)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 70, "deletions": 70}}, {"node": {"oid": "26882c2a6b906a40eaea41886fcf9f197d501a2e", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/dwayner/DmlD\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 3008, "deletions": 1946}}, {"node": {"oid": "395fac9a34d966b6b61fba7fa040d8e824b2e131", "messageHeadline": "Merged PR 5419219: DmlDev RI from GitHub master 2020-11-18", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 4918, "deletions": 1559}}, {"node": {"oid": "6f86c4dbe37aa27c373e58dfacd56ba6f6f4e6c1", "messageHeadline": "Quantize LSTM (#5595)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2995, "deletions": 1909}}, {"node": {"oid": "c2a993e7457192c8ae8e8017aeddf656c89c2dda", "messageHeadline": "Add documentation for OrtArenaCfg for CreateAndRegisterAllocator API.\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 18, "deletions": 7}}, {"node": {"oid": "b3a6ed14d457ab9d77eae2c485d8d749c912ab8f", "messageHeadline": "Prevent saving a model containing fused nodes as we don't have any wa\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 27, "deletions": 0}}, {"node": {"oid": "e8c0f5d0ffd2e7a27ee6c6d768ef29f77df49661", "messageHeadline": "Update the quantization script to support GEMM (transB==1) (#5432)", "author": {"name": "Peichen Xie", "email": "flans39@126.com", "user": {"login": "flans39"}}, "additions": 42, "deletions": 16}}, {"node": {"oid": "f964bb94ba1468104bb353c056aa59335082f511", "messageHeadline": "Add QLinearConv NHWC transformer (#5824)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1360, "deletions": 337}}, {"node": {"oid": "e814c9307a92e60799f03837665c3812374d52f6", "messageHeadline": "Boost Expand cpu operator by multi-threading (#5739)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 365, "deletions": 154}}, {"node": {"oid": "71e7c2b4235a5132e4b9d00a01b18abe3aab71cd", "messageHeadline": "Cache build docker images in container registry. (#5811)", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 429, "deletions": 115}}, {"node": {"oid": "252dbf118273099fb2730f12bffcafbecef1ca00", "messageHeadline": "fix build break (#5835)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "3b5ba1cf7ed8bd8d74fce748416ae01f8687512a", "messageHeadline": "Parallelizing Resize op (#5792)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 90, "deletions": 80}}, {"node": {"oid": "bd236ecc2678d5ea1cb7c2369295d731444ac79b", "messageHeadline": "Switch to unified DirectML 1.4.0 redistributable (#5794)", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 48, "deletions": 60}}, {"node": {"oid": "c84bc25e28214fd8d8eb6ae3a7716c09e92dab5e", "messageHeadline": "Add validation of op registrations (#5817)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 506, "deletions": 295}}, {"node": {"oid": "241b2226a71a7db5ce2c71e681a78ab6bbf3d56b", "messageHeadline": "Update orttraining-linux-gpu-ci-pipeline.yml for Azure Pipelines (#5826)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1a66dfc0f96206be47b1ee295972fec714ba815c", "messageHeadline": "Enable Squeeze Opset 13 for NNAPI (#5717)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 96, "deletions": 17}}, {"node": {"oid": "502f85c8839a7b48d0984a405e6ec810e3e1841b", "messageHeadline": "Merged PR 5411855: Manual pull from onnxruntime upstream GitHub maste\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 11603, "deletions": 4396}}, {"node": {"oid": "7b76b57fc83363acdb02a67834e01fbe1c02443f", "messageHeadline": "Support EPs that compile nodes in a minimal build. (#5776)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1867, "deletions": 506}}, {"node": {"oid": "794e8479eb52139036f3df5dca6c43305c555616", "messageHeadline": "Revert #5805 (#5823)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 7, "deletions": 167}}, {"node": {"oid": "a3f3a63206c0fb43c2114e8811af5306ce452e55", "messageHeadline": "Move OpenVINO specific validation function to somewhere more sensible\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 44, "deletions": 43}}, {"node": {"oid": "732ffd12d2103c7f9bbd8dc77b9a3e0dadafd7ad", "messageHeadline": "DirectML Execution Provider integration 2020-11-13 (#5809)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 62, "deletions": 44}}, {"node": {"oid": "339348bc4664c3b1cb530f02f08639d29bfc39d7", "messageHeadline": "Fix bug in resize IsOpSupported, and add nearest neighbor resize supp\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 50, "deletions": 21}}, {"node": {"oid": "833432d7d18f3c7aa38c69ec8e42a93d9c22430b", "messageHeadline": "Update mysql-connector-java (#5802)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2a6c73cf8cd5627a6bb4bd0038236964e7128f17", "messageHeadline": "Address publishing pipelines failures. (#5806)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "671fa6032778861254fabb357107fe4b3ed1f6df", "messageHeadline": "Enable direct tensorization and detensorization to many buffers in Wi\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1083, "deletions": 268}}, {"node": {"oid": "20ae1ea21fdd1641766f6ecb10260fecb6d283fd", "messageHeadline": "Remerge custom gpu op (#5818)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 37, "deletions": 13}}, {"node": {"oid": "e40df385ba6b8df5e0437bc8cab52fffecab85b0", "messageHeadline": "Skipping even more x86 tests (#5799)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 17, "deletions": 0}}, {"node": {"oid": "89e5b3a24f29ffdca490147ad6aef882cb74e52f", "messageHeadline": "resolve review comments", "author": {"name": "zhijxu", "email": "zhijxu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "89902c251982b5ddc73be67793a4dfc83289c5b8", "messageHeadline": "fix frontend bug.", "author": {"name": "zhijxu", "email": "zhijxu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "c4818d36ed50dc2c8aa05cf67bc98cc0f0748f93", "messageHeadline": "[NNAPI EP] Make NNAPI EP build on non-Android Platform (#5779)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 157, "deletions": 51}}, {"node": {"oid": "5b7dc5aeee636d00ca7e002b2a424a379aace749", "messageHeadline": "fix build failure for ROCm EP (#5816)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ced5b66306cd6884124042cc0d9db714e7444aac", "messageHeadline": "Re-enable multi-tensor-apply for LAMB optimizer", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 3, "deletions": 6}}, {"node": {"oid": "fc614ad050b2ad019ecfa8c80489987eb74fbd20", "messageHeadline": "revert the code change which was based on b4869926", "author": {"name": "Weixing Zhang", "email": "weixing.zhang@microsoft.com", "user": null}, "additions": 455, "deletions": 31}}, {"node": {"oid": "c23fbba463e97487fd44ad93f62e5ddc23f4acaf", "messageHeadline": "Fix reduce pipeline by replacing model (#5813)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 9, "deletions": 13}}, {"node": {"oid": "3269e59b2ce1ec0d4326e08cb8bf9ef10b90c631", "messageHeadline": "Add opset 13 registration for Identity. (#5800)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 17, "deletions": 5}}, {"node": {"oid": "157d1844fb241a43def5ed4e97c6a4ba4cb87af0", "messageHeadline": "Named Dimension Override internals test and experimental API (#5805)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 167, "deletions": 8}}, {"node": {"oid": "262e9ef21dc319a8e7bece1f0d155364b17b019c", "messageHeadline": "Support input dimension swap in Attention op (#5774)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 115, "deletions": 39}}, {"node": {"oid": "dfbf6d78beb12cf72de4aa7b40222b50939aaa69", "messageHeadline": "OpenVino: fix allocation failure on Window for RelWithDebInfo build (\u2026", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 31, "deletions": 11}}, {"node": {"oid": "0c8902cbbe4739914edad9ff6e9dea4845f958a5", "messageHeadline": "Update Gradient Builder of Some Ops for OpSet13 (#5748)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 167, "deletions": 106}}, {"node": {"oid": "1f722863b24c29a6555d5606867755f75dd10acc", "messageHeadline": "Scale Bias post processor for ARM (#5795)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 0, "deletions": 4}}, {"node": {"oid": "435b904f0e8dd845ff4cd9f5efbe7dd66b1d9219", "messageHeadline": "add dnnl gpu engine (#5788)", "author": {"name": "jeyblu", "email": "63478620+jeyblu@users.noreply.github.com", "user": {"login": "jeyblu"}}, "additions": 870, "deletions": 283}}, {"node": {"oid": "0ea998134a1e5cada954e8be4932aba8d11ff067", "messageHeadline": "Skip new x86 tests in ort model tests (#5789)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 12, "deletions": 8}}, {"node": {"oid": "2f35e651353779dd5fd395bc472ce4a0aec6eea5", "messageHeadline": "Add Float16 and BFloat16 support to C# API (#5775)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 451, "deletions": 177}}, {"node": {"oid": "4d517c68a3958b97dcfe8267714ef0af130eced6", "messageHeadline": "Fix reference to old download_e2e_test_data.py script. It was renamed\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "88c3704257e18dbee022390acbb89e4b0b7226d4", "messageHeadline": "Add shape inference for additional ops", "author": {"name": "Alberto Magni", "email": "49027342+alberto-magni@users.noreply.github.com", "user": {"login": "alberto-magni"}}, "additions": 46, "deletions": 1}}, {"node": {"oid": "4e29f4801038e839cacd633118fee5d89882e65c", "messageHeadline": "skip gpt2 test on x86 (#5787)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 1, "deletions": 0}}, {"node": {"oid": "49288de17cde826cd0ea330c39960bc6a64bf618", "messageHeadline": "Fix memory planning issues (#5752)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 23, "deletions": 19}}, {"node": {"oid": "44d3c31200b8bfcb5bdf341b979a1174df19c9d1", "messageHeadline": "Winml_principles_change (#5727)", "author": {"name": "alexzakv", "email": "72228588+alexzakv@users.noreply.github.com", "user": {"login": "alexzakv"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "dc0f7b8f825a2601ad3835836a840b7dceb33b3f", "messageHeadline": "Remove onnxruntime_session_options_config_keys.h from c_api (#5772)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 15, "deletions": 11}}, {"node": {"oid": "54de618c2ee767c8620e787a52b056645adb8bc9", "messageHeadline": "Improve TensorRT engine caching (#5737)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 288, "deletions": 114}}, {"node": {"oid": "2a87108431b6851f5b453434a3d6ca4f5fbcd8d8", "messageHeadline": "SoftmaxCrossEntropyLoss OpSet13. (#5777)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 26, "deletions": 6}}, {"node": {"oid": "b92fc66ea14a310bbc5f31f872a9138cfddb8b21", "messageHeadline": "Support opset-13 specs of controlflow ops (Loop, If) (#5665)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 601, "deletions": 151}}, {"node": {"oid": "07dc25e9390cf37c043c862de3ab41d266cca695", "messageHeadline": "Compute global gradient norm according to 'enable_grad_norm_clip' (#5\u2026", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 200, "deletions": 64}}, {"node": {"oid": "1ae58c960c3618fac8c8cd7e6490cfca08c57ab3", "messageHeadline": "Allow turning off printing of shape when compiled with onnxruntime_DE\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 46, "deletions": 9}}, {"node": {"oid": "5aec34500dd8d2f27c67bb0b1c8ad1e4e0006156", "messageHeadline": "Add megatron transforms for BART (#5521)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 978, "deletions": 215}}, {"node": {"oid": "a14cd6267b186fe5fc5d6b3d3e546a604574bc1c", "messageHeadline": "Support opset-13 specs of softmax family ops (Softmax, LogSoftmax, Ha\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 684, "deletions": 142}}, {"node": {"oid": "e5c8040c52b67b3f4bfd4df792c0bbb99b39f631", "messageHeadline": "Improves performance  of operator Transpose (#5550)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 164, "deletions": 92}}, {"node": {"oid": "a84a058f9e0286e76efd5ad4d14d3634cfc9ed55", "messageHeadline": "[OpenVINO-EP] Enabling Multi Device support (#5740)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 114, "deletions": 28}}, {"node": {"oid": "4207e99be34acfab5c0c1c0dbd4e4b6921a0d992", "messageHeadline": "[NNAPI EP] Move GetCapability independent of ModelBuilder (#5767)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 254, "deletions": 232}}, {"node": {"oid": "d8ace07ad73ac1683ca76b66640f062174df0221", "messageHeadline": "Add CPU send/recv for pipeline (#5315)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 655, "deletions": 127}}, {"node": {"oid": "496fa18c9686dc194d98843f126bac8f64052fb9", "messageHeadline": "fix graph partitioning for nested functions (#5755)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 12, "deletions": 16}}, {"node": {"oid": "bc1768c7f13cde766180a707d132c7b047d1b28a", "messageHeadline": "Stop gradient flowing to the `k` input of TopK (#5762)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 2, "deletions": 1}}, {"node": {"oid": "871af477d7a699d6c6ad46a7a52f1cd106fb6858", "messageHeadline": "Fix outputs of Sequences and Maps exposure. (#5743)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 413, "deletions": 184}}, {"node": {"oid": "1416d12f0b99629138b9da60621a0aadb64a9500", "messageHeadline": "Liqun/merge e2e pipelines (#5702)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 404, "deletions": 8}}, {"node": {"oid": "2ba637c55836c7b2dca554d96bf993970d086606", "messageHeadline": "Implement Scale function for quant gemm (#5632)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 483, "deletions": 387}}, {"node": {"oid": "cca8cd849a507b07571402166eff7ab6fd2d1204", "messageHeadline": "update native build instructions for ACL on Jetson. (#5764)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 10, "deletions": 15}}, {"node": {"oid": "79350a642a01255300f35164592707f966ee88f6", "messageHeadline": "Update install_deps.sh: remove the unnecessary data generating step (\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "0767c4fdfb4de470fe5c3f1bf53b59fac4918280", "messageHeadline": "Fix x86 build break (#5759)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "042365029fa797b81e09bab36932833bde235323", "messageHeadline": "[NNAPI] Split OPBuilder IsOpSupported into a separated class (#5746)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1528, "deletions": 1175}}, {"node": {"oid": "6803e4ab446d75c21e849b050546f298814be8fe", "messageHeadline": "Fix BatchNormalization registrations. (#5750)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 53, "deletions": 6}}, {"node": {"oid": "7d98596bfaac7c5055a1ee8533f2361b09032f5b", "messageHeadline": "Merged PR 5390213: Tile allow 0 in repeats", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c75b7c5c4740688a39c5eab52b3f82b4aad66c17", "messageHeadline": "[CMake] Enable NCCL only when enabling CUDA or ROCm support (#5516)", "author": {"name": "Alberto Magni", "email": "49027342+alberto-magni@users.noreply.github.com", "user": {"login": "alberto-magni"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "48b14b52b8b476a839a307d307e0a39ee4ab63de", "messageHeadline": "Remove Env::Task wrapper around std::function (#5753)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 15, "deletions": 39}}, {"node": {"oid": "2b1ebbc28651512ba459888a975ff2275412ed45", "messageHeadline": "update MCR images table (#5509)", "author": {"name": "leqiao-1", "email": "61653207+leqiao-1@users.noreply.github.com", "user": {"login": "leqiao-1"}}, "additions": 7, "deletions": 8}}, {"node": {"oid": "4c6118eb491ce09d67ef8d3117ee623974b75f8a", "messageHeadline": "Update get_applicable_matrix_reduction() to combine dimensions of 1 w\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 192, "deletions": 88}}, {"node": {"oid": "63b85fc6966d5e6630b7d42253031e6ac71f3653", "messageHeadline": "Fix VS 2017 build break (#5745)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "d59f057db3c893a093c32a830bff039fcb4508e6", "messageHeadline": "enable string for operator Shape (#5742)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "8c74df2068b4b1bdfecabfe19d19a1825c68abab", "messageHeadline": "Add support for string with operator Expand (#5751)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 44, "deletions": 3}}, {"node": {"oid": "4094a09a561726bd60837993b7c20014288c26c0", "messageHeadline": "Merge pull request #5731 from microsoft/snnn/rtti", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "00b18d9dc57d34b2d8361da0c18747ba3d558d20", "messageHeadline": "Update InferenceTest.cs to exclude one more model in x86 mode", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "5e44d25c5a57fe81a2df700d42d99e821eadd576", "messageHeadline": "Support multi-loop parallel sections, use multi-loop sections in GRU \u2026", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 971, "deletions": 330}}, {"node": {"oid": "919c270f3c18915b4897a3ca0014dce3e0546ce8", "messageHeadline": "Increase build timeouts.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "de85638543708869dc9abda760488169cd1ebb55", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1168, "deletions": 976}}, {"node": {"oid": "838bc77f3b12a26f5db27583437dce6ee28fb0c3", "messageHeadline": "Merged PR 5386132: Update 8D BatchNorm", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "2acdc3cd8290c2395c95fcc5f33e719c2abb0fe3", "messageHeadline": "Move GetUseDeterministicCompute() to OpKernelContext to avoid need to\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 36, "deletions": 53}}, {"node": {"oid": "12d39ef4ed2c7069f539daf35fef90c7cce30f0e", "messageHeadline": "Remove onnx backend test filters for updated ops (#5718)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 17, "deletions": 29}}, {"node": {"oid": "bb1af718b524f9aa819f0c5f5fa200e0fe744220", "messageHeadline": "fix build failures due to recent change(858040fa) in CUDA EP (#5736)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 185, "deletions": 0}}, {"node": {"oid": "c0c9ab4d81b6fba0db6a0d97f322f60cdcd33062", "messageHeadline": "Fix kernel registrations for Equal, Greater and Less (#5730)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 930, "deletions": 894}}, {"node": {"oid": "81b2cb97145f2beed6b7003489dea92368d26df4", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 27456, "deletions": 5316}}, {"node": {"oid": "1e13ecabf7a0a3fafbd31d277ca7e035a165bc68", "messageHeadline": "Merged PR 5380534: BatchNormalization failure in autopilot - fix outp\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "2bf5046d4e8a809851c83c5230857a228c3554d9", "messageHeadline": "Add tag types for Ort::Float16_t and Ort:Bfloat16_t structs (#5716)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 129, "deletions": 7}}, {"node": {"oid": "fff85a6a350be201499b22d34db6e55594e087e3", "messageHeadline": "Add GPU kernels for ROCm EP (#5655)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 8235, "deletions": 1006}}, {"node": {"oid": "697e8faa9e681a7e710db0eac8ae9d44f3d2a76a", "messageHeadline": "Skip failing x86 winml tests and update testData environment variable\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 39, "deletions": 24}}, {"node": {"oid": "9ec6da1e276626e2b1217af0cf6c44ff3197dbf3", "messageHeadline": "added missing flag ORT_TENSORRT_DUMP_SUBGRAPHS (#5724)", "author": {"name": "Johannes Bannhofer", "email": "4116408+joba01@users.noreply.github.com", "user": {"login": "joba01"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "6f6dd0b86966de6a8332e69279cef0ade825f99c", "messageHeadline": "added missing flag ORT_TENSORRT_DUMP_SUBGRAPHS (#5724)", "author": {"name": "Johannes Bannhofer", "email": "4116408+joba01@users.noreply.github.com", "user": {"login": "joba01"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "92292de1355085472f311b5f37d66b57587c3fcf", "messageHeadline": "Tensorrt perf tool (#5436)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 889, "deletions": 356}}, {"node": {"oid": "95e6da7957ea901cabc438f53500f16c976bd1f1", "messageHeadline": "Revert saving optimized model as external data (#5690)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 31, "deletions": 21}}, {"node": {"oid": "71f90e08f13c7e39340b0e95fa8876c6f01fdcd2", "messageHeadline": "Nuget packaging no omp (#5666)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 348, "deletions": 1}}, {"node": {"oid": "77b1eea9cfa401add01868060fc089e36b73fc75", "messageHeadline": "Add option to allow quantize_input() use input_qtype for initializers\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "f666c3d7d7ab40f04dd37c50a9d25e23c9496d51", "messageHeadline": "update jetson build instructions (#5725)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "24016a517bfa4d9dbc85768e045c95718221dc94", "messageHeadline": "Prepacking in Gemm with merged logic for Matmul and Gemm on PackingB.\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 275, "deletions": 130}}, {"node": {"oid": "479ed740efa214666a93c45b382fc92e0242fd94", "messageHeadline": "Add link to survey to README (#5685)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "d6f9cc181d2077f34b954192e2b545f462144544", "messageHeadline": "Modify logic to determine OV Version (#5701)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 13, "deletions": 5}}, {"node": {"oid": "d1d82065b9db9127b32a083f71a07b42d348d44c", "messageHeadline": "[Java] Fixes an error allocating large direct byte buffers during Onn\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "28197b146070ca882aa34264795b53b092dd91e3", "messageHeadline": "Register opset13 flatten, LRN for cuda. (#5694)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 47, "deletions": 14}}, {"node": {"oid": "11fe6834710938140d000691369de47c68daa6ed", "messageHeadline": "Partition full graph one execution provider at a time (#5635)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 126, "deletions": 81}}, {"node": {"oid": "858040faaaccf79d661853e823a28829f483e165", "messageHeadline": "Implement reduce_matrix_columns() to optimize ReduceSum (#5639)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 967, "deletions": 474}}, {"node": {"oid": "c46515cd56c93586864491bc32c7975eb6bea254", "messageHeadline": "[TensorRT EP] Remove cudaDeviceSynchronize and use cudaAllocator for \u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 32, "deletions": 36}}, {"node": {"oid": "fd9d0c4ee019f57e0d4009f8279ebb52c616d106", "messageHeadline": "Remove redundant const_cast (#5705)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9e68e984230df9a780372353dff090e605a968d4", "messageHeadline": "Add static CRT DLLs to Nuget package (#5661)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 188, "deletions": 35}}, {"node": {"oid": "ff23083de2fb1a1f20e79566b422127fc3cb75a8", "messageHeadline": "Unbreak microbenchmark build (#5710)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 41, "deletions": 15}}, {"node": {"oid": "5c4543e194be0b7ff753b24e5d5c32949fbde5c3", "messageHeadline": "Calibrate float tensor only (#5704)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "2127a229d7f9640f171d1df005844680a2e92ed1", "messageHeadline": "The IndexedSubGraph is used to create the Function body, but after th\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 40, "deletions": 48}}, {"node": {"oid": "941e3a69f92628188cb4d7c2da3ef6d14a1d94cc", "messageHeadline": "Test a build break fix (#5706)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "6d8e81cb08ead9115bcf8d8936c56a60141cf51b", "messageHeadline": "Update Squeeze, Unsqueeze, Split and ReduceSum kernel for Opset13 (#5\u2026", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 729, "deletions": 266}}, {"node": {"oid": "830f567be83be732f8721ffac296ba1ce20327c3", "messageHeadline": "Add C API Guidelines document (#5686)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 87, "deletions": 1}}, {"node": {"oid": "8bae883d3ef3d776825568babaff309c5f43a932", "messageHeadline": "User/alexzak/win ml principles (#5453)", "author": {"name": "alexzakv", "email": "72228588+alexzakv@users.noreply.github.com", "user": {"login": "alexzakv"}}, "additions": 36, "deletions": 5}}, {"node": {"oid": "62a99824cbd1cce01533b15865f62973390a93e1", "messageHeadline": "Wezuo/priority in nodedef (#5692)", "author": {"name": "wezuo", "email": "49965641+wezuo@users.noreply.github.com", "user": {"login": "wezuo"}}, "additions": 35, "deletions": 22}}, {"node": {"oid": "e49f7a8b71d64596f74ab03be6960ab576b02a16", "messageHeadline": "Fix build error due to unused variable (#5698)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "0b9f7bb1b0d56d7fffccd0557525d6c03229d137", "messageHeadline": "Update InferenceTest.cs", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "0445473dc148da8b06002af7a2490c2eafd3def3", "messageHeadline": "Add ssd to x86_disabled_tests", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "a2b551ff08aba07cee64bd282f8afc71d54fcf20", "messageHeadline": "Add runtime options for NNAPI EP (#5576)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 77, "deletions": 38}}, {"node": {"oid": "2ad7bcb766f96d271a89c1892364e4ca57df3e30", "messageHeadline": "NNAPI add opset version check (#5687)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 72, "deletions": 11}}, {"node": {"oid": "07bd4ef47008065f6abe0ac1766515a5c58c07c6", "messageHeadline": "Upgrade optional implementation to https://github.com/martinmoene/opt\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 100, "deletions": 71}}, {"node": {"oid": "67d7e3967de77dc088729cb557954907c7aada7b", "messageHeadline": "Disable some model tests", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "b6eeadf420ed93103fea5b0b95395b32700b5d89", "messageHeadline": "Enable OpenVino build on Arm64 platform (#5682)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "c9f44276da0919c4c7e7409210142beb61c8ab11", "messageHeadline": "Add ability to filter GraphViewer using IndexedSubGraph.  (#5614)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 294, "deletions": 39}}, {"node": {"oid": "357a51c75cab5b030d5b16dd63bf689357069c11", "messageHeadline": "Update python packaging pipeline's docker image (#5680)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "db9c1308a5c6f159d49f9466e5104c59375eb57c", "messageHeadline": "Fix Resize kernel registration (#5677)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 42, "deletions": 20}}, {"node": {"oid": "28f1e328982796bad5fcf42c87c86a2d9d7b9eb0", "messageHeadline": "Loosen tolerance of CudaKernelTest.ReduceSum_MidTensor, allow test ra\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 108, "deletions": 54}}, {"node": {"oid": "a028ca41ecc1df40b49b6fa8bdf11287816bea16", "messageHeadline": "Optimize flaubert (#5651)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 48, "deletions": 4}}, {"node": {"oid": "9b010963b7ecd7b03099ffbfe1c509b0cca4e7ef", "messageHeadline": "Turn off peak memory logging and fix memory pattern generation bug. (\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 24, "deletions": 13}}, {"node": {"oid": "5d66cf017cc4ea417ee68fa07a1a7d3359d8e276", "messageHeadline": "Register Clip for OpSet13 (#5671)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 17, "deletions": 3}}, {"node": {"oid": "8856c2595bb53a5480c49ba31b442e3ec7d7d2e3", "messageHeadline": "Sync the two IDs in OrtMemoryInfo when calling ctor (#5663)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "4936e10e22f4fc003288dd08ea9a70207700139d", "messageHeadline": "Disable some model tests (#5664)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 2}}, {"node": {"oid": "182d9c48e4eae385388ffb35990d76d26d277265", "messageHeadline": "Merge u8u8/u8s8 QLinearConv implementations (#5662)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 343, "deletions": 296}}, {"node": {"oid": "c875fe09193cb9c6bbe4d572bffab62b30bebb12", "messageHeadline": "Add option to dump activations on all ranks (#5455)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "87e1063e19fd57eb955d04f021e7b80377cbfc0b", "messageHeadline": "Revert \"Update Squeeze, Unsqueeze, Split and ReduceSum kernel for Ops\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 266, "deletions": 717}}, {"node": {"oid": "2c025306038ac2222b3d0f2f6dca1f27a0cb1e79", "messageHeadline": "Bert Model Profiling Tool (#5654)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 237, "deletions": 1}}, {"node": {"oid": "1495f737ca995fe6f5312f8688708c94962eb7cf", "messageHeadline": "Use cudaMemsetAsync and add checks on CUDA calls.", "author": {"name": "Jesse Benson", "email": "benson.jesse@gmail.com", "user": {"login": "jessebenson"}}, "additions": 27, "deletions": 31}}, {"node": {"oid": "db63c5d10fcccd28ba5a6edb3c3798eab63c6948", "messageHeadline": "Update Squeeze, Unsqueeze, Split and ReduceSum kernel for Opset13 (#5\u2026", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 717, "deletions": 266}}, {"node": {"oid": "5b44982971ba33dfab40d1c60e40ed1147970be3", "messageHeadline": "Change the OrtCustomOp invocation as a constant. (#5506)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 26, "deletions": 25}}, {"node": {"oid": "ff538b8d3a1c6264f70e7b4ab1ecf480cc655e6a", "messageHeadline": "Minor fixes in BERT Inference notebook (#5637)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "1cca90368014699b49d6ae29413d571ee1b22265", "messageHeadline": "update onnx commit id (#5594)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 28, "deletions": 28}}, {"node": {"oid": "f2168cef296ac180b3c451fb5e199446fb69b185", "messageHeadline": "Misc. cleanup. (#5659)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 4, "deletions": 9}}, {"node": {"oid": "9af0d485241d3f037b94ebf876e457f396b6e583", "messageHeadline": "Memory planner and pattern generation enhancements. (#4443)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 519, "deletions": 89}}, {"node": {"oid": "d98062da0ce1ffb804325396cde10a41933cd154", "messageHeadline": "[OpenVINO-EP] Hetero support (#5627)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 291, "deletions": 58}}, {"node": {"oid": "d9293f38e60b2a41dc7551c1b32557a89a567ded", "messageHeadline": "Revert \"Custom Op on GPU (#5620)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 29}}, {"node": {"oid": "7948a4b0bcc16708ba95d4bfce2be064fc022040", "messageHeadline": "Revert \"add header (#5648)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "32bf6390ad1697438f46db9a0189625a6bb0ee22", "messageHeadline": "Some fixes to symbolic shape inference (#5642)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 27, "deletions": 7}}, {"node": {"oid": "7a80a4b526c79ebdf82adae99ea104ebdd1ca614", "messageHeadline": "Support more C# APIs  (#5608)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 299, "deletions": 16}}, {"node": {"oid": "17bce6f07ec786c429807f51a3be2dcecea0e63c", "messageHeadline": "Implement Im2colNd NHWC and related qlinearconv logic for u8s8. (#5612)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 377, "deletions": 196}}, {"node": {"oid": "d7f3baed181f6b8891fcc0cea1d4cd3403779b36", "messageHeadline": "add header (#5648)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "3e71e8bd7e60b11eb4b32afacf852089090d2572", "messageHeadline": "Revert \"[CUDA EP] remove per-thread allocator (#5415)\" (#5647)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 63, "deletions": 26}}, {"node": {"oid": "2c63196600c191dbc419db9bb3d04deddc18e35e", "messageHeadline": "Custom Op on GPU (#5620)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 29, "deletions": 12}}, {"node": {"oid": "aa38893afb299003754bb59f66dec4bd99fd1a18", "messageHeadline": "[OpenVINO-EP] Add Dockerfile with C# API bindings (#5633)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 96, "deletions": 7}}, {"node": {"oid": "aec4cb489e49a63fc8523ee3d22e406c2af9c22b", "messageHeadline": "ROCm EP for AMD GPU (#5480)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 5415, "deletions": 85}}, {"node": {"oid": "742ffb860c7e4528521f851c4e74995dfe91fe61", "messageHeadline": "Allow Kernels refer to some attribute data directly in the protobuf (\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 125, "deletions": 15}}, {"node": {"oid": "1fa1c51544ad60e7779dd809e5ae7b00b6bdd67a", "messageHeadline": "bug fix for name of gradient constant (#5626)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "b4869926d338daaa1011a76e22f3e059ca79cf04", "messageHeadline": "[CUDA EP] remove per-thread allocator (#5415)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 26, "deletions": 63}}, {"node": {"oid": "2e1fa3ccb75a6441ee0a8a6da92c396daac1816f", "messageHeadline": "Fix GeluRecompute for 2 inputs case. (#5573)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 68, "deletions": 2}}, {"node": {"oid": "b85e7a19eabc2ba92545a8a62ac7a2cadfe531ff", "messageHeadline": "isalnum is not defined - include cctype (#5623)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "e6956be40cf063d6bdc69485ad3be8ddb84ffa87", "messageHeadline": "Publish no-openmp python packages to test pypi (#5610)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 208, "deletions": 2}}, {"node": {"oid": "b68e98e0b0c13bde4954885cc1c998b104686a03", "messageHeadline": "optimize QLinearConv depthwise convolutions (#5605)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 286, "deletions": 81}}, {"node": {"oid": "1d87831c6e8719eb3d2b07c8650ab0d3bf157177", "messageHeadline": "Merged PR 5344477: Disable GPU timeouts in DML EP command queue creation", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5129b4d5bc8426e5fc23047010dbf3fe0d14a99c", "messageHeadline": "batch size tests (#5508)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 213, "deletions": 93}}, {"node": {"oid": "50582abe937fbae4a1426c34b368e877ee6e7088", "messageHeadline": "Fix IS_ANDROID Issue  (#5599)", "author": {"name": "Rohith_Kvsp", "email": "phani.rohith@gmail.com", "user": {"login": "Rohithkvsp"}}, "additions": 7, "deletions": 14}}, {"node": {"oid": "bbfd914d72d0ccfbf0658de5fc46574fcd9c733b", "messageHeadline": "Skip new model test additions (#5611)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 37, "deletions": 11}}, {"node": {"oid": "27c6d1eeb245cb79f18d9d6e36e7f4aaba06eea0", "messageHeadline": "move variable declaration to avoid unused variable error (#5603)", "author": {"name": "Juliana Franco", "email": "jupvfranco@gmail.com", "user": {"login": "jupvfranco"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0dbf3e889358cce6dded935c5623b2cce707b690", "messageHeadline": "enable arena for arm64 (#5613)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "5e8952ef89b52aec7a0646840cf6fe693cedfd48", "messageHeadline": "ThreadPool clean up : mm_pause in loops, correctly spin-then-wait, an\u2026", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 113, "deletions": 58}}, {"node": {"oid": "92662659ba2da60f224aa6d608afba8044a39348", "messageHeadline": "Liqun/remove number matching (#5606)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 126, "deletions": 230}}, {"node": {"oid": "e90b6f06d13aa64812e3c55bc43240f9111de29c", "messageHeadline": "Factor out IAllocator so that it can be shared with shared providers \u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 95, "deletions": 238}}, {"node": {"oid": "e5b0d192f4f9910d6188f1d3c5d8f6ce3caf37a7", "messageHeadline": "pin transformers dependence to sentencepiece==0.1.92 due to ci fail (\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ddf83d1acef3c21dcbe44814870b76696038d1fc", "messageHeadline": "Maajid/multi threading 2 (#5568)", "author": {"name": "Maajid khan", "email": "n.maajidkhan@gmail.com", "user": {"login": "MaajidKhan"}}, "additions": 153, "deletions": 68}}, {"node": {"oid": "b851973f227863acce798f1b0e597a8b3c06bc38", "messageHeadline": "pipeline_worker_pool_.JoinAll() should be called in pipeline code pat\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 3, "deletions": 8}}, {"node": {"oid": "6f824c25e525616448b5efaa731f9721ab65a84e", "messageHeadline": "Dropout op elimination - enable for ORT training (#5588)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 122, "deletions": 0}}, {"node": {"oid": "3433576fd39bb6451fb520e208a2f34a07ba4c7b", "messageHeadline": "Support for Sparse Initializers (#5540)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 911, "deletions": 313}}, {"node": {"oid": "30cdc74bc00bb0859c99a0699b703e2889c317df", "messageHeadline": "Enable prepacking in subgraph (#5433)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 218, "deletions": 61}}, {"node": {"oid": "564da960cefaba2c715f3cae58bebbd45f1b9d14", "messageHeadline": "Fix nuphar docker file build break", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6c310858e3d221a1d5c6ec8b0ab0afcf2438005c", "messageHeadline": "Support opset-13 Resize kernels (#5575)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1027, "deletions": 920}}, {"node": {"oid": "5bcb5f5a3d25fa0c33b3909a5f9b1e8c57617bf4", "messageHeadline": "MLAS: Add support for AVXVNNI (#5592)", "author": {"name": "Ramakrishnan Sivakumar", "email": "63471807+ramkrishna0116@users.noreply.github.com", "user": {"login": "ramkrishna0116"}}, "additions": 1676, "deletions": 0}}, {"node": {"oid": "e380fd3c6beb1628ac1ee54d49ec0999832cc58b", "messageHeadline": "Merged PR 5334334: Fix asserts and failure in GraphKernelHelper.cpp", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 49, "deletions": 40}}, {"node": {"oid": "694a4d6413eb2b9450835ccdb5bd2dc4c5ce2238", "messageHeadline": "Add more loggings for GradientBuilder (#5556)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 51, "deletions": 27}}, {"node": {"oid": "68fe722691a291e622ac969dffb847b292003fa7", "messageHeadline": "GatherGrad optimization (#5524)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 964, "deletions": 501}}, {"node": {"oid": "8224718f8f57b7c0079d399a7138d6053c876b23", "messageHeadline": "Enable CommonSubexpressionElimination in training. (#5504)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 54, "deletions": 0}}, {"node": {"oid": "44773c60e3bdad278cfae41e652c4bb06605c15f", "messageHeadline": "Add  a CUDA based IOBinding test (#5572)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 131, "deletions": 8}}, {"node": {"oid": "f4cee22b9bec2eac51f8a16c0d5aed78ef6031a6", "messageHeadline": "Handle -inf in ReduceSumLogExp, fix regression introduced in PR #5370\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 120, "deletions": 2}}, {"node": {"oid": "502f67ba58ac3aa45b65423fa5f11f1193a69a79", "messageHeadline": "MLAS: implement u8x8 GEMM for aarch32 (#5580)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 761, "deletions": 17}}, {"node": {"oid": "b2da700e4d953239833e40f9a1b39b15936cc6dd", "messageHeadline": "Allow Upper case letters in RHS of einsum equations. (#5569)", "author": {"name": "Andrew McDowell", "email": "35566934+AndrewMcDowell@users.noreply.github.com", "user": {"login": "AndrewMcDowell"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "51af108af5d620bf6372c3eae766e66c7f2d412a", "messageHeadline": "Support older version of slice in reshape fusion (#5574)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 82, "deletions": 9}}, {"node": {"oid": "860cb22260058056bb9f9d528820ef91d7782178", "messageHeadline": "Bug fix for C API (#5520)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 15, "deletions": 11}}, {"node": {"oid": "3f3b202e36d8fe5aa8d5455e8564d959170d4c13", "messageHeadline": "Optimize GatherElements further, add threshold for parallelizing Scal\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 131, "deletions": 53}}, {"node": {"oid": "49ec73e939e49e8e9e659d1c0e6703188e5901f1", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 591, "deletions": 178}}, {"node": {"oid": "3f06286154df52d5900e306b5a6dd19b53194ed1", "messageHeadline": "Add Flatten support for NNAPI (#5545)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 338, "deletions": 100}}, {"node": {"oid": "7da5949279577ee21fa09ed6484833dfe10f127d", "messageHeadline": "NVTX label change (#5562)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "20bc83400b6a0a5f6a27c5d24f5742eb63335115", "messageHeadline": "ACL/ArmNN update (#5515)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 252, "deletions": 77}}, {"node": {"oid": "00d749b3c975f3eb3c30c1a66e363556d6d65560", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2673, "deletions": 951}}, {"node": {"oid": "98538580c89ab5c9eef99b53496c5cb679a0eaf4", "messageHeadline": "give more tolerance to DirectML runs (#5564)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 7, "deletions": 0}}, {"node": {"oid": "1f304fbee77f38fa2919b4e895f03c3e7b722433", "messageHeadline": "Attention with past and no unidirectional mask (#5557)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 208, "deletions": 38}}, {"node": {"oid": "0a9b83a31392fc6b81b6d492c8488bfae016383a", "messageHeadline": "Add zero test (#5476)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 106, "deletions": 2}}, {"node": {"oid": "6d35be215fe9e43363d267cf995c747fbeef4d96", "messageHeadline": "Add `--skip_tests` to example command line as the included ops are be\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d220c9f9504131132a1e2a314a1d4ec4e38676cb", "messageHeadline": "Resolve crash in MatMul optimization (#5551)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 39, "deletions": 0}}, {"node": {"oid": "5802fe1699ec7bf3f6b1de370d6c5190f0af0ff6", "messageHeadline": "Remove MKLML build config (#5559)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 18, "deletions": 513}}, {"node": {"oid": "82c7a9756e9f67916e2282202a6f70cee319ffc8", "messageHeadline": "Fix shared provider unload crash (#5553)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 130, "deletions": 82}}, {"node": {"oid": "4291c57322a4e709f51a44416da747dc715e6ad9", "messageHeadline": "[C# and Python APIs] Expose knobs to enable/disable platform telemetr\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 163, "deletions": 25}}, {"node": {"oid": "df226110263ce5e569af2975a781aba097c98037", "messageHeadline": "Update ONNX commit (#5487)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 357, "deletions": 150}}, {"node": {"oid": "b48f596a91db787242ecec74c338730e445ec37b", "messageHeadline": "GatherElementsGrad CPU Kernel and TopKGrad CPU/CUDA Kernel (#5511)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 320, "deletions": 59}}, {"node": {"oid": "6c2162e97add696e9f7d0377dba322a4d1fe05cd", "messageHeadline": "Fix quantization of Conv1D with bias (#5491)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 18, "deletions": 26}}, {"node": {"oid": "1038f9cc8bb269512777d25d3f76e9fe869da40b", "messageHeadline": "Optimize GatherElements and Scaler. (#5543)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 85, "deletions": 61}}, {"node": {"oid": "2f4fc83231d8ca75fc0d92b569064f5386d5027f", "messageHeadline": "Add NVTX profiling range around kernel computation. (#5542)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 18, "deletions": 13}}, {"node": {"oid": "45483dcf1f63bf21767cc18468b2e51cb1e1f661", "messageHeadline": "Add QLinearConv for activations=u8, weights=s8 (#5510)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1324, "deletions": 59}}, {"node": {"oid": "280cdf31f582c0f964073055ce5543227d3ff33e", "messageHeadline": "Revert \"Fix shared provider unload crash (#5523)\" (#5547)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 81, "deletions": 124}}, {"node": {"oid": "21e02effc0a313be9278a2de2334b3c92c6a318f", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 898, "deletions": 780}}, {"node": {"oid": "66c8a441e08ebee0fc2494c4a597b1f0bc177d08", "messageHeadline": "Improves ReduceSum performance by removing transposition. (#5370)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 608, "deletions": 574}}, {"node": {"oid": "682898ae2ba4690e7caf126aac3dd7c7b3b0e9c8", "messageHeadline": "Add #include for std::tolower. Fixes VS2017 build error. (#5544)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "67968441e0e3e670326ea20af973afd3318d48dc", "messageHeadline": "GatherND - add Cuda support for int64 on opset 12 (#5531)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "610676293ebfe4800e3c2b6a5da1c57e54bb1280", "messageHeadline": "Fix shared provider unload crash (#5523)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 124, "deletions": 81}}, {"node": {"oid": "4b2942365628d8f2926a906e89768d95ce257fd1", "messageHeadline": "Re-enable custom op shared library test for debug builds (#5475)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 48, "deletions": 11}}, {"node": {"oid": "0298b9734e498b4c2569241afdf46e19949c5775", "messageHeadline": "Save in EndTraining only if in last rank  (#5500)", "author": {"name": "Juliana Franco", "email": "jupvfranco@gmail.com", "user": {"login": "jupvfranco"}}, "additions": 20, "deletions": 12}}, {"node": {"oid": "a3d2bc36beb06fe28336690e52c8fcaeb93896cd", "messageHeadline": "Fix script name in doco (#5530)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6ad70d73710ff3b98a1c3815d21351b89a08e3e3", "messageHeadline": "[Doc] ONNX_Runtime_Server_Usage fix proto uri (#5345)", "author": {"name": "Thien Bui", "email": "thien@canva.com", "user": {"login": "ducthienbui97"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1e4b259d28ceafb6cd52f0743418002222db9f89", "messageHeadline": "Updating EP docs with Onnxruntime API calls (#5503)", "author": {"name": "Olivia Jain", "email": "oljain@microsoft.com", "user": {"login": "oliviajain"}}, "additions": 40, "deletions": 96}}, {"node": {"oid": "0b59004666595b3d8aec1ab8af27075fc868a3cb", "messageHeadline": "Add fallback function implementation for DivGrad (#5518)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 36, "deletions": 3}}, {"node": {"oid": "a355281b99e2be533f7d626abfa03059a4c1e744", "messageHeadline": "Add alternate IsSupportedOptypeVersionAndString signature (#5529)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "bc84aeb5a8fd99ac745aa5ca4dbfcfeaccd54267", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 574, "deletions": 530}}, {"node": {"oid": "e1a54c4090de70784574f0f1157b828198fc01a2", "messageHeadline": "Symbolic shape inference: fix a bug in shape merge (#5519)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "eda9fd566e08ea654656ee628443fdd4df6b2ca0", "messageHeadline": "Update tar-stream and prebuild-install versions (#5479)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 571, "deletions": 527}}, {"node": {"oid": "ec0dd8edbf5645917357f086dc4cb60596252675", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1605, "deletions": 580}}, {"node": {"oid": "ad94a1dd6de37b3c3db8ba9a377ff7f3549520ea", "messageHeadline": "Add opset 13 registrations for Identity, IsNaN, NonZero, GatherND and\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 172, "deletions": 78}}, {"node": {"oid": "f207f0bf5e172c4465318e4bacc81b314e4ae9bf", "messageHeadline": "Add WinML Model testing (#5417)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 536, "deletions": 1}}, {"node": {"oid": "b991ee4c696bc3b3352f740ce2a23cd0f7b1258e", "messageHeadline": "Cleanup NNAPI code (#5505)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 118, "deletions": 141}}, {"node": {"oid": "6f65e2ad2ce07f57541c97b78b00d2a916c46fd0", "messageHeadline": "Mark the dX and dB outputs of ConvGrad as OpSchema::Optional. (#5462)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "64f6d856e4f06e4f4e54c257695a761e6e5c7f2a", "messageHeadline": "Add FlattenGrad and test. (#5461)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 22, "deletions": 0}}, {"node": {"oid": "88f6523baf05a5e6faf583c914b10edb53579b2b", "messageHeadline": "Add type inference for BroadcastGradientArgs (#5501)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 7, "deletions": 1}}, {"node": {"oid": "7da7e07909094fad33ede674a6c6a78eb48b2881", "messageHeadline": "Cleanup some test infrastructure (#5484)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 298, "deletions": 335}}, {"node": {"oid": "645d97858930a015a2cf2fbedc81df0bf36eeb67", "messageHeadline": "Sunghcho/denormals (#5391)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 450, "deletions": 22}}, {"node": {"oid": "c1602ff0285db98774732e7a8fe30590fef373ae", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1300, "deletions": 395}}, {"node": {"oid": "915d47535394a72a050bd29e91766de8ba8538ff", "messageHeadline": "Android CI update (#5474)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 25, "deletions": 9}}, {"node": {"oid": "6d2a30eae30f09e48584bd48d11ea0605d91c76b", "messageHeadline": "[OPENVINO-EP] 2021.1 Release  (#5431)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 1275, "deletions": 386}}, {"node": {"oid": "88ebad4cfe1e1fc587e64f43e0f0f1d8b1002cd7", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2627, "deletions": 1160}}, {"node": {"oid": "2b6b3a2ee6fdef5b4670a4017f889004981fa05a", "messageHeadline": "Add GetProfilingStartTimeNs() to Python/C# APIs (#5280)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 102, "deletions": 8}}, {"node": {"oid": "1514509fd702032171febc6d242ebdbde1cbaab0", "messageHeadline": "Update protobuf submodule url (#5477)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "44248d9646b8f006e47148eac5b251a91d1bdbe3", "messageHeadline": "opset13 kernel registration (Transpose, Tile, ScatterND, ScatterEleme\u2026", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 134, "deletions": 28}}, {"node": {"oid": "fabe02ddc2e802e75fa90db003779ba961b90523", "messageHeadline": "Don't change global FPU state during round-half-to-even (#5376)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "67315d8ae0212c7184b3647602369dd68e1789bc", "messageHeadline": "Optimize openai-gpt/albert model and add fusion test (#5466)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 147, "deletions": 27}}, {"node": {"oid": "5544391e7984e8591d7aa1d0c4d6c208d4073228", "messageHeadline": "Fix linking of MLAS unit test lib on platforms where libatomic is req\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "8e9afe19448e9d302097de1323cc130b20fb80b8", "messageHeadline": "Add long type support for SplitToSequence operator (#5367)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "e01d152464b1c229bb513391e331858ddb317ad8", "messageHeadline": "Add OpSet kernel registrations as part of opset 13 support (#5465)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 146, "deletions": 79}}, {"node": {"oid": "6e6147fb7533a902c79a55911ced1ef54b00be79", "messageHeadline": "Use correct protoc tool file name for C# builds (#5429)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b12824fa7ace324c9d7328c8fb7bff724f92b141", "messageHeadline": "add telemetry event for nodejs binding (#5463)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 21, "deletions": 4}}, {"node": {"oid": "ce5465d5f3e97b01b868a6107fdbce84175010d3", "messageHeadline": "[NNAPI EP] Add Resize and Clip support (#5427)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 631, "deletions": 114}}, {"node": {"oid": "c444b9d76a9fda3775026ed3a4120ae1a6caaeb3", "messageHeadline": "Add CUDA option to run copy in default stream (#5445)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 93, "deletions": 37}}, {"node": {"oid": "80d36eab8698cbfd52b66bce11537aa529b1d56e", "messageHeadline": "enable the onnxruntime shared library test on iOS (#5443)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 52, "deletions": 34}}, {"node": {"oid": "913116e64ea20af0600e39f6284aaaad60335da5", "messageHeadline": "bump ops version to opset13 (#5456)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 262, "deletions": 154}}, {"node": {"oid": "05b1c02d32c2a1919744583ef9a40314cbcb3603", "messageHeadline": "Fix commands in README.md. (#5459)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "60dbd8a1e54eb75bab47848fb4072aa3d99bde94", "messageHeadline": "Update maximum batch size for UT; Include recompute modes  (#5444)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 30, "deletions": 12}}, {"node": {"oid": "dbc626dcbe3e2e17529f0696759b3b7bc95918e2", "messageHeadline": "Add ExpGrad registration and test. (#5438)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 30, "deletions": 0}}, {"node": {"oid": "2a018cc23521d812ea2d781a539834ee276b9afc", "messageHeadline": "revert contrib op version bump and deprecation of TransposeMatMul (#5\u2026", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 38, "deletions": 55}}, {"node": {"oid": "20c47ce91c0fe6ce01ad37f4b8386a19a699753b", "messageHeadline": "Simplified layer norm changes (#5028)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 739, "deletions": 503}}, {"node": {"oid": "ed60e0fe392fcb1769653d44f3e6fcff1e488970", "messageHeadline": "Fix BUILD.md environment variable name typo. (#5402)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5e48c0fd6c8231e9dfd6a5a52abdcc8e9b765013", "messageHeadline": "Register opset13 ops: Dropout, Flatten, LRN, MeanVarianceNormalizatio\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 168, "deletions": 90}}, {"node": {"oid": "17544fc9720f76c7e750b10a92f6429b76b9ee8f", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 651, "deletions": 430}}, {"node": {"oid": "186f0668b0ef5557db1827f90e3ad777265afd1e", "messageHeadline": "update onnx-tensorrt submodule (#5442)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b9f90e297e0f2c7199cd8215ece7e37f852207d7", "messageHeadline": "Support sharing of initializers between session via the Python API (#\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 93, "deletions": 53}}, {"node": {"oid": "6132e1f6ae690ddb44053f015da37786390d0167", "messageHeadline": "Shared providers - fix logging plus cleanup (#5406)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 155, "deletions": 174}}, {"node": {"oid": "6cba42e9420d62b8df996ac64fdb891b40790068", "messageHeadline": "Avoid inserting other CUDA calls in-between NCCL Send's and Recv's (#\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 42, "deletions": 19}}, {"node": {"oid": "dbe7e6623b62445415e355da878572285f6ed689", "messageHeadline": "only use/import pytest if needed (by enable_training) (#5437)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "9642f1448e1fefcef37411343ceda154ae7eec6a", "messageHeadline": "Add OpSet 13 Registrations (#5426)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 158, "deletions": 76}}, {"node": {"oid": "3a9a1a4ef198dc81eed2e9378ccd75994941545a", "messageHeadline": "Fix registration for GatherGrad (#5382)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 5, "deletions": 3}}, {"node": {"oid": "1cceefc7d4c48851b3d7ff9030cc8c27d54a8d0b", "messageHeadline": "use run_orttraining_test_orttrainer_frontend_separately to work aroun\u2026", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 31, "deletions": 1}}, {"node": {"oid": "a92ccbe1bc015d0b679b08f0c623d4be164ff7b3", "messageHeadline": "Various armv7 related fixes (#5394)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 164, "deletions": 102}}, {"node": {"oid": "8aa61aaa47372e23a7aa639ee090905d0b281ffe", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 4125, "deletions": 1461}}, {"node": {"oid": "b99eaa99cdf3cc4403564ebce2a210eb2db5dd90", "messageHeadline": "Prepacking MatMulInteger (#5403)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 112, "deletions": 78}}, {"node": {"oid": "621fdb44e5c91d7a9e2d440df30e57167ca159a2", "messageHeadline": "Fixes #4688, remove CPUAllocator in TreeEnsemble (#5375)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 15, "deletions": 14}}, {"node": {"oid": "d4507e93315eca38622024ac5715f8c621e64401", "messageHeadline": "Use relative path for HEADER_SEARCH_PATHS (#5412)", "author": {"name": "Keizo Fujiwara", "email": "keizo@pyull.com", "user": {"login": "keizof"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "90f976d0603eedc02f88e9ec0deea207305e6edd", "messageHeadline": "Some improvements on transformers tool (#5383)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 60, "deletions": 11}}, {"node": {"oid": "fab7f799a7fac819d334e9f337204caea5f067b7", "messageHeadline": "MLAS: fix ARM64 + VS2017 build break (#5423)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "8a632a903f983780a9818b1bbf2fef9e88d610c3", "messageHeadline": "Remove unused imports from Python tests. (#5405)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 3, "deletions": 31}}, {"node": {"oid": "15696b8fceed4b403be966ef5caca60adeb0a4f9", "messageHeadline": "bump version to 1.5.2 (#5420)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 12, "deletions": 6}}, {"node": {"oid": "498f94668da2443df191e11bc39442208a778016", "messageHeadline": "Keep all_finite tensor on CPU when using PyTorch Frontend (#5371)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 6, "deletions": 1}}, {"node": {"oid": "c2c78399ee751e5a0d3c642f580af9b192012cfa", "messageHeadline": "Include config keys header file in the release packages for Linux and\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "09aef240d61a6ae15f8d9556a1913b94588cf973", "messageHeadline": "Skip running onnx tests in python mac os pipeline (#5416)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "83ead3e2eb7ad01b6e8889cabab9fab21dc18fc1", "messageHeadline": "Fix com ptr refcount (#5404)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 7, "deletions": 5}}, {"node": {"oid": "b04cf2d2290f7550fc5aa633e118845dd11c1fa5", "messageHeadline": "Update ORT to 1.5.1 in Bert Quantization Notebook (#5396)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 108, "deletions": 100}}, {"node": {"oid": "132ab2230de72463c0cda250a81d216960524998", "messageHeadline": "Updated with image for creating the onnxruntime pkg (#5400)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "9684e1b5a8b13bce7a91b00041ab0229ca678c2f", "messageHeadline": "Add doco for pre-requisites to be able to cross compile for Android o\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "81332238715710794ee02239cf50add20ff264b7", "messageHeadline": "clear cudaDelayLoadedLibs since delayload is disabled (#5386)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "8ee2b08325ef7b4c53fe935a1d9bcfa94f2fb813", "messageHeadline": "Allow benchmark different threads (#5390)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 115, "deletions": 87}}, {"node": {"oid": "094384781ea0caa3931061609ca90a84b6a0b64c", "messageHeadline": "Add --use_external_data_format in convert_to_onnx.py (#5393)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 67, "deletions": 26}}, {"node": {"oid": "594744545747b1eb16da40439df94304eec1423c", "messageHeadline": "Add flatbuffers verifier for ORT format buffer (#5378)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 567, "deletions": 521}}, {"node": {"oid": "deb708d3b12138d87135fabf71da616526087279", "messageHeadline": "Move flatbuffers to 1.12 release (#5392)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "6f54113a1bcea4e643de844ee33e4914145915ef", "messageHeadline": "Support OrtValue binding in Python to enable interesting IOBinding sc\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 834, "deletions": 277}}, {"node": {"oid": "0122e890d9ecdd20a89055b8e0244f43c5f18004", "messageHeadline": "MLAS: implement u8x8 GEMM for ARM64 (#5380)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1934, "deletions": 245}}, {"node": {"oid": "b4934b0016dae533020bfc4767aab078acb9163c", "messageHeadline": "Mitigate pybind11 build break using Xcode 12 on macOS (#5381)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 20, "deletions": 3}}, {"node": {"oid": "10f1902d90a516fd6123c6ae7f2f712152d99f63", "messageHeadline": "Update code snippet in README.md", "author": {"name": "Kaarthik Sivashanmugam", "email": "skaarthik@users.noreply.github.com", "user": {"login": "skaarthik"}}, "additions": 19, "deletions": 10}}, {"node": {"oid": "773992c7d4d9262f2b4501166afd31d9b2fd0e69", "messageHeadline": "Liqun/bert pretrain tb (#5377)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 145, "deletions": 35}}, {"node": {"oid": "b5caa7cb12146e2d7a422bf009d2eafeedffc3ec", "messageHeadline": "Updated docs: Execution Provider overview (#5328)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 64, "deletions": 0}}, {"node": {"oid": "1a3bfd7f06c760f098026620304de9f4065eca08", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 377, "deletions": 181}}, {"node": {"oid": "323c4dfe0245300f1b35cc937da48652daad5d81", "messageHeadline": "Adding an option for cudnn conv algorithms. (#5159)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 156, "deletions": 57}}, {"node": {"oid": "a0b8218f9a601d3a292625cc60993d3db57e1a5a", "messageHeadline": "Amdmigraphx update to rocm3.7 (#5362)", "author": {"name": "Shucai Xiao", "email": "shucai.xiao@amd.com", "user": {"login": "scxiao"}}, "additions": 9, "deletions": 11}}, {"node": {"oid": "24f99b3be86f68fa16ba8ac97328dae2e2a244bf", "messageHeadline": "Support OuterStride for QGemm when MLAS_SUPPORTS_GEMM_U8X8 undefined \u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 77, "deletions": 23}}, {"node": {"oid": "668ab049175974ba109b7045ca3a7b21823116b8", "messageHeadline": "rename all TransposeMatMul nodes to FusedMatMul (#5373)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "891175d5d66187798a6ea08c59e620fb1c1ca4c1", "messageHeadline": "Merged PR 5253310: Fix 0-sized dimension broadcasting", "author": {"name": "Patrice Vignola", "email": "pavignol@microsoft.com", "user": null}, "additions": 7, "deletions": 0}}, {"node": {"oid": "4e3a420aa7695242650e8a7df843acc13bfbe041", "messageHeadline": "Use single thread when pipeline is not enabled in TrainingRunner (#4265)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 134, "deletions": 89}}, {"node": {"oid": "3165d8045b0f94cafdd741d6f27bc076ed712895", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1317, "deletions": 342}}, {"node": {"oid": "c20fcf26ebc0eeda41b6fd4e54d87623030ead91", "messageHeadline": "Onnx GPU runtime fails to fallback to CPU when GPU is not available/b\u2026", "author": {"name": "Vlad Burlik", "email": "vladbph@users.noreply.github.com", "user": {"login": "vladbph"}}, "additions": 17, "deletions": 7}}, {"node": {"oid": "4721729fdc2195191670f8c9459e8b76e80e630f", "messageHeadline": "Enable iOS CI pipeline (#5360)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 16, "deletions": 5}}, {"node": {"oid": "9df07908568b385b9d1ba72c576f2cadeb0f6219", "messageHeadline": "Update linux minimal CI to report Android mininal baseline binary siz\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 84, "deletions": 15}}, {"node": {"oid": "5bd7241839acd74ae424c002ed8696f0be836d73", "messageHeadline": "Raise output mismatch error in ort_test_dir_utils.py (#5364)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "f5e4c0ea04acb6e5dcdc804a7e8935e12aad553d", "messageHeadline": "Fix benchmark_gpt2 model verification (#5343)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 83, "deletions": 24}}, {"node": {"oid": "6e4949e235e780be12093a104e74ca205e4bb792", "messageHeadline": "javadoc warning fix (#5332)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "06cd81d79199ab826a239d894be38dc7b9830abb", "messageHeadline": "Support trilinear sampling in Resize CPU and CUDA kernels (#5300)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 578, "deletions": 89}}, {"node": {"oid": "e71668f92c1ec7be50fdcd05ce4eebf59dce2af8", "messageHeadline": "Expose recompute configs to the frontend (#5318)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 262, "deletions": 51}}, {"node": {"oid": "e33de20861e86a0625d4da7d68fbad35c4f281ec", "messageHeadline": "Update gpt2 notebook for int8 quantization (#5346)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 241, "deletions": 134}}, {"node": {"oid": "ce49cfa67ce6dcdfd5b56f71ede792e7db9f698a", "messageHeadline": "add support for configurable build dir when building nuget packages (\u2026", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 26, "deletions": 13}}, {"node": {"oid": "bb3c8f970bc85862ba672f01e3cd7af1bd57064e", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 6939, "deletions": 104}}, {"node": {"oid": "f265834c2c81adefa8230fd6c03520b7f930cc00", "messageHeadline": "Exclude GPT2_LM_HEAD from OpenVino's model test list (#5356)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "1612934f72bc015c8f295f0e9f3d41f19bbed4ef", "messageHeadline": "Allow protobuf format of input data for performance test (#5323)", "author": {"name": "Sunghoon", "email": "35605090+hanbitmyths@users.noreply.github.com", "user": {"login": "hanbitmyths"}}, "additions": 6093, "deletions": 48}}, {"node": {"oid": "e8b9aa1f2933f59a6418a208192d1f65b628e05b", "messageHeadline": "fix quantization of EmbeddingLayerNorm (#5321)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "7495dc167a359872c92cff99f8ebb516a943f13f", "messageHeadline": "Symbolic shape inference: fix a bug in auto_merge when broadcasting (\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 5, "deletions": 3}}, {"node": {"oid": "caed6c264c869dcffd181a5e9628ae1962c914a4", "messageHeadline": "Add tf2pytorch wrapper in transformers tool (#5316)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 193, "deletions": 9}}, {"node": {"oid": "d62873a331a7e9de23f383f21426d69df86ebed0", "messageHeadline": "Docker image release build updates (#5326)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 9, "deletions": 10}}, {"node": {"oid": "fe5021349160af75f1481b6a88d44a7686e13330", "messageHeadline": "Liqun/bert pretrain2 (#5327)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 576, "deletions": 4}}, {"node": {"oid": "1cad3e322ebc33f03f95343626dd70eeeff77c78", "messageHeadline": "typo in contributing.md (#5340)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2098d621a6fa47a7dd21b996fd6832d6961a5ed0", "messageHeadline": "Make some string optional for save to/load from flatbuffers (#5331)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 59, "deletions": 26}}, {"node": {"oid": "a343b21a09040869409f82c5150679abc3f36aca", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 820, "deletions": 491}}, {"node": {"oid": "383b1e207c772f6afd0344e4a7dd73e68f63a5be", "messageHeadline": "Fix bug in the Resize operator kernels (#5303)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 117, "deletions": 42}}, {"node": {"oid": "3f00b8db8f45ae92509d2128561703ef62cb910f", "messageHeadline": "move all experimental ops to version 1 of ms domain (#5287)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 677, "deletions": 429}}, {"node": {"oid": "2c32309e2cfe92153bd913b6a8d7ec3edef39653", "messageHeadline": "Update dockerfiles/README.md onnxruntime-training image tags. (#5333)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "37445d11980eab3f310c5e491e671a48d6c6d95b", "messageHeadline": "Update Bert Perf Script (#5339)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "8d4740b39ca78f3430867d5d0b8aa3e1a6da22cf", "messageHeadline": "Add some log for the GetFileLength function (#5330)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "cb57c100e609c0e543319d360e1697dacff298af", "messageHeadline": "Doc updates for 1.5 (#5302)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 17, "deletions": 15}}, {"node": {"oid": "e237b9d7ede0356ff8d2e44442046063355dee80", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 55, "deletions": 8}}, {"node": {"oid": "69dbaaa015beff0e83cbfdc52363d1a1c9cdb51b", "messageHeadline": "Add additional test cases to check for leaks in thread pool creation \u2026", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 37, "deletions": 0}}, {"node": {"oid": "1a12f510fc0c18519e545f04cf5358d65a1ea633", "messageHeadline": "Support T5 benchmarking in transformers tool (#5133)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 18, "deletions": 8}}, {"node": {"oid": "c443330e981de6ff4c3b389d9617507935f8975f", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2072, "deletions": 116}}, {"node": {"oid": "9ec1ed42a809170b87474f5822c4557101812399", "messageHeadline": "Enable BiasDropoutFusion for CUDA EP only (#5324)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ed102e9d880614953277ab80580b2f524e1b379c", "messageHeadline": "Add iOS test pipeline and a sample app. (#5298)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 1857, "deletions": 23}}, {"node": {"oid": "f07059ccc03bfc17510dbe2c5f7376afa3b46521", "messageHeadline": "Add weight prepacking to LSTM kernel (#5305)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 213, "deletions": 91}}, {"node": {"oid": "5af3630a2992a64f64ba3be2c3ec43d1b51c844d", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 1606, "deletions": 562}}, {"node": {"oid": "11c194ce290eb6309c44c1ecf1991a84f26dbf51", "messageHeadline": "Minor fix for ComputeBroadcastBackwardAxesDynamic; Fix for GradientGr\u2026", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 8, "deletions": 9}}, {"node": {"oid": "24d8b1bf428a9bd157d632ed132c4bc4e6f3f60c", "messageHeadline": "to skip an unstable test to unblock release (#5314)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "cb8309763251899cf9385e6dbb24c66b2f66b6e8", "messageHeadline": "Cosmetic change in non tensor tests (#5317)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 8, "deletions": 7}}, {"node": {"oid": "1ff3b2d5b8d7de2059516e41eb2465a90d73762b", "messageHeadline": "Add ability to generate multiple test dirs so that different input mi\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 4}}, {"node": {"oid": "eae2473dc15b29124162c5640bda203186d35d1c", "messageHeadline": "Scale Op for ReduceMeanGrad. (#5191)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 424, "deletions": 22}}, {"node": {"oid": "506060dc374d35e5e7154c721cd8a5b4362d8417", "messageHeadline": "Remove Useless Cast from Contiguous Cast Nodes (#5204)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 81, "deletions": 19}}, {"node": {"oid": "d45d68fdd41b45111aa6d07fc3633e9ddb5583a7", "messageHeadline": "Fix a memory leak in our testing code (#5312)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "3693f9121885fc0df572dbecbe30b1e7cb437dfc", "messageHeadline": "Update doc to be explicit about backwards compatibility. (#5309)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "b18a8bc74f074d27ed92597c9f4ccbfbf657ab3c", "messageHeadline": "Transpose kernel fix for illegal memory access error (#5294)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "1a04b8f8b7160a6c12ea9e94d98ca16500d1febd", "messageHeadline": "Add valgrind support to our cmake files (#5296)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 19, "deletions": 9}}, {"node": {"oid": "fec890a09aa58cc7d7260ee4fb9ed8a9eb24579a", "messageHeadline": "fix build break (#5306)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "507f5bf5f6281f2ca8f6ede908afad64bff0890f", "messageHeadline": "Update test calibrate script (#5185)", "author": {"name": "RRRachelllll555", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 229, "deletions": 94}}, {"node": {"oid": "d9ecc0cebf8752893d4cd4e547341e390dc29b01", "messageHeadline": "add bert loss legacy back (#5224)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 171, "deletions": 0}}, {"node": {"oid": "16d35266ab9e060c5866a0cb629e37b50015130e", "messageHeadline": "add install targets for ep shared libs (#5286)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 13, "deletions": 0}}, {"node": {"oid": "3a3f26f38e049eb0e8b9fbb97281a6e5dd82074f", "messageHeadline": "Move ort flatbuffers helper functions and value info r/w functions in\u2026", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 628, "deletions": 394}}, {"node": {"oid": "098f6b9069d821a095105b548180b35725f6fbbe", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1520, "deletions": 818}}, {"node": {"oid": "17f1178c2e77e1aeda0bf37f04097a677d6439f0", "messageHeadline": "Downgrade GCC (#5269)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 274, "deletions": 205}}, {"node": {"oid": "b03fb82ab77b0bb4ed6f2a49a7aea4adcff6b538", "messageHeadline": "Transformer layer-wise Recompute (#4526)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 1246, "deletions": 613}}, {"node": {"oid": "c2c30285cd7d75724e8a7832005d34e576d7c377", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2137, "deletions": 1310}}, {"node": {"oid": "b6e71200eb37da9bcdec03b703fd3984a8728198", "messageHeadline": "Add additional tutorial links (#5272)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 26, "deletions": 7}}, {"node": {"oid": "89742411ecfee064e23c3ca7eb3be48b904ebc58", "messageHeadline": "Insert telemetry template into GPU build, add telemry build switches.\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "ebeeff22dd7774e6a1a72fa328f8698a74cf96ba", "messageHeadline": "Update PyTorch TransformerModel sample (#5275)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 191, "deletions": 2}}, {"node": {"oid": "71b52ad5de1e99eb051df000acfadea8551a6dfa", "messageHeadline": "Fix inbox telemetry  (#5265)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 9, "deletions": 2}}, {"node": {"oid": "b49ff6151e4dc080cd9fe8200fd2672e9bbba51c", "messageHeadline": "Workaround issue with VS2017 compiler. (#5279)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 21, "deletions": 2}}, {"node": {"oid": "5a71819be6f738f2d334ee707634cb621ab5d843", "messageHeadline": "Symbolic shape inference: fix a case for concat (#5277)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 17, "deletions": 8}}, {"node": {"oid": "4ed31ca21406fdaa37d2491d755bd52765b4ce9b", "messageHeadline": "Combine custom logger global threadpools (#4857)", "author": {"name": "Josh Bradley", "email": "jgbradley1@gmail.com", "user": {"login": "jgbradley1"}}, "additions": 57, "deletions": 26}}, {"node": {"oid": "6ad39819c20c278a4c3ec20ee838373008ed7ec7", "messageHeadline": "Update DirectML Nuget to 1.3.0 (#5274)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "a4cb00b91e9b86522355a6be0aead9c23f01d5df", "messageHeadline": "Merge pull request #5273 from microsoft/user/dwayner/CmakeLinkerOptFlags", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "6d5b93b805ee181d8b96662257784960b87828d4", "messageHeadline": "Synchronize training dependency versions between Docker image and Pyt\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 63, "deletions": 105}}, {"node": {"oid": "56862f4022322f349e10d41db168b5df4f315b2f", "messageHeadline": "Add way to disable additional linker opt flags", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "16220f38480b094d34e94afd76f48afa814543ab", "messageHeadline": "Add FusedMatMul contrib op (#5213)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 343, "deletions": 290}}, {"node": {"oid": "fe7e7bfe6017212074998b4454629848ea26966e", "messageHeadline": "Update build.md for building using Xcode 12 on Mac (#5256)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 19, "deletions": 7}}, {"node": {"oid": "61ba5b501a18ce6ac074cf39132a30fd05487398", "messageHeadline": "Fix bug in the back to back quantization of matmul and conv (#5264)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "b5a6a8e84762ecbf13f890c8486c6e8e946fbbc7", "messageHeadline": "remove implicit linking of tensorrt and dnnl ep shared libs (#5262)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 4}}, {"node": {"oid": "6ea66b43db0a78a4766f741e0de94ac09cb73632", "messageHeadline": "ORT DirectML EP for Iron release, ONNX 1.5 (part 2) (#5263)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 41, "deletions": 18}}, {"node": {"oid": "75d994f1940444022bc22127c87f56636df884a6", "messageHeadline": "Handle zero norm values in LpNormalization CPU kernel (#5251)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 39, "deletions": 2}}, {"node": {"oid": "d26c71f55cdf512ea495c738b94ac69fdf045616", "messageHeadline": "[java] Fixing the buffer semantics. (#5223)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 156, "deletions": 140}}, {"node": {"oid": "c52561d0441f2640cdb78f19239e800fde05362f", "messageHeadline": "Rework broadcasting setup to decrease binary size. (#5227)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1150, "deletions": 695}}, {"node": {"oid": "43faf9e388fe1790d5130c29b5b45364b9c57ff5", "messageHeadline": "Disable a few tests that run too long(1 hour) in debug mode (#5257)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "3bbce69185bb157014952fcc0a0f7172a2ad2e8a", "messageHeadline": "bump version to 1.5.1 (#5258)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 10, "deletions": 9}}, {"node": {"oid": "59e69bf35ba1d05d14ae605b944b87b34d95e240", "messageHeadline": "Handle missing initializers in allocation planner to fix crashes with\u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 20, "deletions": 4}}, {"node": {"oid": "682ac275fb3266c5f5d5b016e25515b70a03d0b8", "messageHeadline": "Merged PR 5201369: Remove copy of initializers added in DMLXP refactor", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 25, "deletions": 12}}, {"node": {"oid": "9d57bb4ba772668e9666e6a3ae2084fa404e35b5", "messageHeadline": "Add way to disable additional linker opt flags", "author": {"name": "Justin Stoecker", "email": "justoeck@microsoft.com", "user": {"login": "jstoecker"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "ed5665758ce81e4b22101c2f2c725378ab0205a9", "messageHeadline": "Merged PR 5195856: Fix broken cases of zero size tensors in Cast/Reduce", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 16, "deletions": 6}}, {"node": {"oid": "898531f502f745e83f1ec29093b33f65867d12b4", "messageHeadline": "Fix reshape fusion crash (#5252)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "e30530d9ea214f2e540a3d21ab444d1b633c0de4", "messageHeadline": "Add java API for AddSessionConfigEntry (#5241)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 55, "deletions": 0}}, {"node": {"oid": "8dceebda0ed9214b5c43fd69c57abf5f8d720274", "messageHeadline": "[Training/Python] Add option to enable symbolic shape inference (#5107)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 140, "deletions": 74}}, {"node": {"oid": "14f250a4d076346846bf25f899fb4f2301309b99", "messageHeadline": "Update BUILD.md training dependency info. (#5240)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 13}}, {"node": {"oid": "d957dbebea93aafc0a63c4ee60df4aaeca6d6711", "messageHeadline": "Fix possible ios build break after update to Xcode 12 (#5246)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "417929b049829c44bcd59c0d0eae7ae6c71ab111", "messageHeadline": "jobs timeout ..", "author": {"name": "suffian khan", "email": "sukha@OrtTrainingDev1.af05slrtruoetgaxwwjv5nsq5e.px.internal.cloudapp.net", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a6eb90472cccef82d71003dc320ae1160fd0dcc4", "messageHeadline": "try fix error on code coverage ci build", "author": {"name": "suffian khan", "email": "sukha@OrtTrainingDev1.af05slrtruoetgaxwwjv5nsq5e.px.internal.cloudapp.net", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "14786432157806392c75e9564d18c448d2cf6092", "messageHeadline": "Place Shape's output in CPU memory (#5245)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "038192bdb28d9b6a4012e36546590936f47a652d", "messageHeadline": " Place shape related compute nodes in CPU  (#4940)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 184, "deletions": 56}}, {"node": {"oid": "0cb09374c68a2eb6851b742a2cb45e59886ab018", "messageHeadline": "Update BUILD.md for CUDA versions (#5239)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3147bc00c3ecebc936979d09bd63bcc60d00b992", "messageHeadline": "update TensorRT docs (#5238)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 17, "deletions": 10}}, {"node": {"oid": "55e4b5d302e74e580d24e76d8d237d6c6a6f66e0", "messageHeadline": "add pipeline distributed training test (#5222)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 69, "deletions": 3}}, {"node": {"oid": "84c222126c8a4f90281b1ec81e12c627fe19ed37", "messageHeadline": "Deprecate testMNISTTrainingAndTestingOpset10 (#4927)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "974b9bfc09126641cced37b285cff05590590c26", "messageHeadline": "Allow sharing of initializers between sessions. (#5092)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 403, "deletions": 47}}, {"node": {"oid": "e0719a1073d521a11a9fbcb553cbc4d8c8141763", "messageHeadline": "Revert to using release SafeInt repo now that it supports a build wit\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "e9671e93f0ac0c6149a85a38ced2d7825701e8c5", "messageHeadline": "Fix TransposeScaleMatMul and MatMulScaleFusion issues (#5230)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 168, "deletions": 65}}, {"node": {"oid": "65740deb105d686988a1a9ffdff0f97a3fa1b7a5", "messageHeadline": "Fix a bug in EmbedLayerNorm fusion (#5150)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 52, "deletions": 56}}, {"node": {"oid": "aefb2cc49b56aa680509213ceaa8d252ebee6052", "messageHeadline": "Create profile for all dynamic shape input tensors (#5229)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 10, "deletions": 14}}, {"node": {"oid": "cd663d58f55476f591dddafbb338b9ac6059d97f", "messageHeadline": "Fix WinML warnings (#5228)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 44, "deletions": 32}}, {"node": {"oid": "78a29aebbcbd0c3b6dab734f221e0f3bf1e24c97", "messageHeadline": "[ORT Mobile] ORT Minimal E2E CI (#5200)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 122, "deletions": 16}}, {"node": {"oid": "8ee4e8226eafd76ee3bd29b302ca4f6b03395576", "messageHeadline": "Preserve relative order of the results and the tests. (#5225)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 22, "deletions": 19}}, {"node": {"oid": "b49f6a5e2c98bb981bb670d0eb09e1fd3d28bf14", "messageHeadline": "using GPU_WARP_SIZE to make kernel portable between AMD and Nvidia GP\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 12, "deletions": 17}}, {"node": {"oid": "84589c7e0521a2e01e953aa9ee8d90ea13db9eb2", "messageHeadline": "Fuse softmax(a + b) in case of simple broadcast (#4937)", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 1509, "deletions": 4}}, {"node": {"oid": "e0b49844e924527b5de6932c45882618c7c9fb24", "messageHeadline": "Provide option to let layernorm stash mean/var as fp32 or bfloat16 (#\u2026", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 31, "deletions": 5}}, {"node": {"oid": "a90ab12589ce3d4a4323889a7bb77ea67b5e356a", "messageHeadline": "Refactor onnx_test_runner (#5169)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1403, "deletions": 1187}}, {"node": {"oid": "13318ab0d4985b4526fc9b6bc5ce7f3679763186", "messageHeadline": "Remove invalid install line (#5219)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "a632dd2d3bc2a708761d95194e3b694577257d02", "messageHeadline": "Amdmigraphx improvements (#5158)", "author": {"name": "Shucai Xiao", "email": "shucai.xiao@amd.com", "user": {"login": "scxiao"}}, "additions": 214, "deletions": 256}}, {"node": {"oid": "f91248e0ccdeb56b971f253cdf054050afdb8a7d", "messageHeadline": "remove curand_generator_ related code since it is not used. (#5220)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 0, "deletions": 19}}, {"node": {"oid": "ce3b67e0cdff354d14315562d2eea50baf764e82", "messageHeadline": "[Python] Move symbolic_shape_infer from nuphar to tools (#5162)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 59, "deletions": 44}}, {"node": {"oid": "f7c1e518109e9e6089976f6bf34fd344dc6ad62b", "messageHeadline": "Remove shape inference and fix save large model(>2g) issue (#5210)", "author": {"name": "RRRachelllll555", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 33, "deletions": 43}}, {"node": {"oid": "c46a480306c48170996a6b02c3bd3ab62fc52653", "messageHeadline": "Update conversion script and process to simplify creating ORT format \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 124, "deletions": 120}}, {"node": {"oid": "1b61dfaf696ef87c99aa0c9330559d57cc95b206", "messageHeadline": "fix _WIN32 (#5218)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "f5df96256c9af7415083a267b07c25228493438b", "messageHeadline": "Fix order of returned values in quantize_weight_per_channel (#5205)", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f37e1292a1d8cbf52e969a2b72d357ece594e793", "messageHeadline": "--shm-size=1024m to fix nccl shared memory issue (#5214)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "8156e0dd109ac67011556a34ca390d616f45d58b", "messageHeadline": "[ORT Mobile] Some updates to iOS/Android build settings (#5184)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 66, "deletions": 46}}, {"node": {"oid": "86981571122b6a513416fc395dcf5930256d580c", "messageHeadline": "NCHWc optimizer fixes for quantized models (#5203)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 83, "deletions": 48}}, {"node": {"oid": "d535894297d482820f541a0863a42084c5cd23df", "messageHeadline": "Add API to allow configuration of the global thread pools. (#5199)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 72, "deletions": 3}}, {"node": {"oid": "e01e0b2e40fd4a8c01ad35b450bd4fd9efa8c13c", "messageHeadline": "Fix softmax_warp_backward math when is_log_softmax = True and registe\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 146, "deletions": 45}}, {"node": {"oid": "584638e5d35128b3523990f51fa232ce6d145955", "messageHeadline": "Corrects doc typos and formatting (#5201)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "cd0386b6497e2933d891de1e2f45268f479bd37b", "messageHeadline": "MaxPool versioning in quantization tools. (#5194)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "b11c1063461ee15bb8e2e584620c51393bbf9b29", "messageHeadline": "Remove almost all of the reinterpret_casts from the provider shared A\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 234, "deletions": 239}}, {"node": {"oid": "c37472a1aa9ce5c9215af83f6d92948390aae7d6", "messageHeadline": "Mixed Precision Transformer and Gradient Builder Refactor (#4892)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 678, "deletions": 689}}, {"node": {"oid": "f3f119a945ae856020d1572b4df211075355afdf", "messageHeadline": "Use onecore umbrella lib in onecore builds (#5182)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 11, "deletions": 14}}, {"node": {"oid": "1a2e289d2d954057ce8d707f7047c49bcfbbe5bd", "messageHeadline": "Fix nuget build (#5163)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 35, "deletions": 39}}, {"node": {"oid": "e6f85f338e7843d75fb6e8997bd32eaba651780a", "messageHeadline": "Refactor TensorAt, prepare for release (#5180)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 40, "deletions": 22}}, {"node": {"oid": "a20f8037f6d1e512d33e4e1d6a8598e3e71f2889", "messageHeadline": "Install ssh in builder image, fix segfault in TrainingRunnerTest.Basi\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "400ac8556567ca3f4b7628a9b687a82678b1d3c5", "messageHeadline": "Improve error message for FE model export checking (#5156)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 13, "deletions": 8}}, {"node": {"oid": "965e2b095d7fe63496ff932a0eceb3f1afc029db", "messageHeadline": "Update MCR CUDA docker image to 10.2 (#5181)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "79e27d937aba6fe4224eadcb765e92c8e6d15300", "messageHeadline": "MLAS: add sgemm weight prepacking (#5183)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 831, "deletions": 223}}, {"node": {"oid": "3afc2bfa73b5a217403a89ab9403b93c01e25fbe", "messageHeadline": "Remove mutable arguments from symbolic_shape_infer (#5166)", "author": {"name": "Oliver Rausch", "email": "Oliverrausch99@gmail.com", "user": {"login": "orausch"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "7f3aa3a1631c5224878af9922c5177de8ef544b2", "messageHeadline": "Add GetStartTime() for profiler to get private profiling_start_time_ \u2026", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 85, "deletions": 3}}, {"node": {"oid": "a0a435abc66b0311df1508a6048cd7460d31e5bd", "messageHeadline": "Add sympy==1.1.1 to Linux docker image (#5177)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 22}}, {"node": {"oid": "0752fd7425ba510b50c4b9cd1d3d02b69acdf553", "messageHeadline": "change version number from 1.4.0 to 1.5.0 (#5178)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 12, "deletions": 6}}, {"node": {"oid": "9f526f45ac065131a5bb28c7e511b073decb24ea", "messageHeadline": "TensorRT Perf Tool (#4900)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 2113, "deletions": 0}}, {"node": {"oid": "ef496d36ea793d5ca7acce75b35cdf4e580c0bf0", "messageHeadline": "Build: Add missing EXCLUDE_FROM_ALL to ONNX submodule (#5161)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "de6e3fb61dfd053ddd486b8b52efd38cb97c627e", "messageHeadline": "Reduce IOS shared library size by symbol file. (#5171)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "8fa427b264e00fb9e8a2f2b03ca78e808f0d0609", "messageHeadline": "Ryanunderhill/backout 5014 (#5167)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 142, "deletions": 541}}, {"node": {"oid": "089789c135732d139038aef1f78ee8bfa87e5698", "messageHeadline": "Revert change to disable support for loading ORT format models in the\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 32, "deletions": 37}}, {"node": {"oid": "c0d7c8bc44185e02485ec787b7ba73e1435508fc", "messageHeadline": "Add docs indicating that the onnxruntime engine from other distributi\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 32, "deletions": 14}}, {"node": {"oid": "1dde215d96fe5b20f118997cd5806c7324681637", "messageHeadline": "promote cuda version on packacking pipelines (#5154)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "3068a835f16892d8122d71d212f2460629362ce5", "messageHeadline": "Fix quantization of 1-D conv with bias (#5157)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 18, "deletions": 9}}, {"node": {"oid": "82b25e1731f62bd62bd203c03a8b4a1b6dbc45a0", "messageHeadline": "Fix datasize call in calibrate (#5110)", "author": {"name": "Andrei Shadrikov", "email": "notvuvko@gmail.com", "user": {"login": "vuvko"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f7edf0aa57c82315642be3fa248fb1175455d2b7", "messageHeadline": "[OpenVINO-EP] Enable EP config options for VPU hardware (#5119)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 305, "deletions": 130}}, {"node": {"oid": "d45e49dd2bfe45cfa2c98e56ac7879bdd185b40f", "messageHeadline": "Add LeakyRelu and Sigmoid QLinear Quantization support (#5116)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 43, "deletions": 1}}, {"node": {"oid": "8946d212bfe8d5cb41d0b22ae238961fa7b788c3", "messageHeadline": "Remove the dependency on CUDA SDk's version.txt (#5155)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 74}}, {"node": {"oid": "20b2f45b24f1516c29ddde16100461353cb74f13", "messageHeadline": "Support per-channel quantization of weight tensor (#5057)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 148, "deletions": 126}}, {"node": {"oid": "2a456d16c035fa406de8f49bef2f18bd68b19a67", "messageHeadline": "Enable onnxruntime iOS shared library build. (#5148)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 15, "deletions": 2}}, {"node": {"oid": "cc3212f9d52ea2bd1711870bef46378c3724ea54", "messageHeadline": "Add fp16 pow kernel (#5016)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 66, "deletions": 29}}, {"node": {"oid": "1d6a21fd0870aa25dc701cdd94089a2999bb9fd0", "messageHeadline": "[TensorRT] Add slightly faster hash computation for `vector<int>` (#5\u2026", "author": {"name": "Moshe David", "email": "mosdav165@gmail.com", "user": {"login": "mosdav"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0c7e9fb52a50c5ae02a5dcc5e422d0b82590d930", "messageHeadline": "changes to ensure compilation issues in windows is fixed by disabling\u2026", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9392aa2f64171eb6734e62e9a844b2e6a4618353", "messageHeadline": "Promote Cuda version to 10.2 for windows pipelines (#5138)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 8, "deletions": 139}}, {"node": {"oid": "6fcd99f6ed27115c771bb256ba43190142e9371c", "messageHeadline": "Some minor updates for ORT mobile (#5146)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 21, "deletions": 13}}, {"node": {"oid": "e5892dd6f3442d83e132985ab8f5ba80df36a195", "messageHeadline": "Change the version check of ort format save/load (#5140)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 31, "deletions": 8}}, {"node": {"oid": "323a1ba8a4a9da404577f50503f116c755bfa711", "messageHeadline": "Add option to exclude support for loading ORT format models in full b\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 109, "deletions": 36}}, {"node": {"oid": "5302fe407931292b9fff2178e523cc31039c996f", "messageHeadline": "A fix in load_pretrained_model() (#5137)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "849bb8653fe53d5787e193eadf081b9c36192b66", "messageHeadline": "Exclude a few python tests for the DML EP (#5135)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 42, "deletions": 14}}, {"node": {"oid": "698eccf15e9a7d17ea8c86673da1fbabdade315a", "messageHeadline": "Add iOS build instruction (#5125)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 42, "deletions": 1}}, {"node": {"oid": "7511021e0eef126d4f115e87560270c6fa38d673", "messageHeadline": "Save Gpt2 test data (#5132)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 67, "deletions": 18}}, {"node": {"oid": "120e3cda74bc33fcc94b191a3e6d0b2df937dd0c", "messageHeadline": "fix path (#5131)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "92a8c650ada604d3b485f33c6b2c8323f89b68a3", "messageHeadline": "[Debuggability] Add feature to ORTTrainer Frontend (#5124)", "author": {"name": "Rayan-Krishnan", "email": "rayankrishnan@live.com", "user": {"login": "rayankrish"}}, "additions": 105, "deletions": 1}}, {"node": {"oid": "89509f256a945eb5ae34fc6aa4fa89c6ab803821", "messageHeadline": "Not fuse SkipLayerNorm when add has initializer input (#5123)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "cd56ab197c924388d167d02da89370d5fc38ea0c", "messageHeadline": "csharp build documentation (#5121)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 22, "deletions": 34}}, {"node": {"oid": "15d431f39b6a4a7cd22d19a93a700db629c823aa", "messageHeadline": "Bump node-fetch from 2.6.0 to 2.6.1 in /nodejs", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "ccfbc56388695f150c9395046d1d186d95babf53", "messageHeadline": "Handle dummy mask in Attention operators (#5108)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 79, "deletions": 10}}, {"node": {"oid": "c794c88ae0fa5a12d237cb502393928c7315520a", "messageHeadline": "Solve name conflict in TensorRT engine caching (#5128)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 9, "deletions": 12}}, {"node": {"oid": "51f3d3af720ab0cfd077703491653d6a900812a6", "messageHeadline": "Enable onnxruntime_perf_test for ORT minimal build (#5126)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 32, "deletions": 9}}, {"node": {"oid": "59ee8ffb17ffce7d0bce9692d8d58555df909ff9", "messageHeadline": "Remove SparseTensor support from minimal build. (#5114)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 191, "deletions": 78}}, {"node": {"oid": "879751f3b7fe7f46b0cb5c1109604338c25ee807", "messageHeadline": "Support Tensorflow benchmarking and onnx export in transformers tool \u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 252, "deletions": 77}}, {"node": {"oid": "c5efb0085de3222b3d6b96c784758aae05d947f9", "messageHeadline": "Update Linux GPU build pipelines to CUDA 10.2 (#5120)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 16, "deletions": 208}}, {"node": {"oid": "a8557b3f0fbf45a2646b38769d5027dbd676c3f3", "messageHeadline": "skip tests when model opset > released opset (#5096)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 51, "deletions": 31}}, {"node": {"oid": "782ccff207fa23c43d812f0035e8d621907a2b64", "messageHeadline": "Add dll probe path so that the right DirectML.dll is loaded while run\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 31, "deletions": 4}}, {"node": {"oid": "5618b9dddc3f19cd301b3f7888b1ec42421a308d", "messageHeadline": "Use CMake built-in function to compare NCCL version (#5118)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 4, "deletions": 7}}, {"node": {"oid": "c5d4ae04018cdbd2c8be041cc67aae68cff76084", "messageHeadline": "Add transformers tools to python package (#5090)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 61, "deletions": 36}}, {"node": {"oid": "61051396e883e9496b37ffcd84b18386856daad0", "messageHeadline": "[TensorRT] Align naming convention and remove redundant code (#5094)", "author": {"name": "Moshe David", "email": "mosdav165@gmail.com", "user": {"login": "mosdav"}}, "additions": 12, "deletions": 14}}, {"node": {"oid": "fae5915d766ede7e14645e22911f958cc090f2bf", "messageHeadline": "CMake fixes/tweaks for minimal builds and MinSizeRel builds (#5112)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 38, "deletions": 23}}, {"node": {"oid": "a5530358c9236f20e2ad3d6ce968219b21426874", "messageHeadline": "Fix a path problem in Dockerfile.manylinux2014_cuda10_2 (#5106)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "47554a04229b8b04e22047e028c50a6679e359ce", "messageHeadline": "Disable some tests (#5103)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "3207de276c1868b938358faac4e58a844379134e", "messageHeadline": "Remove IDeviceAllocator class as it doesn't extend IAllocator in any \u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 112, "deletions": 128}}, {"node": {"oid": "5b6643cefb5477cbd124b874e0f1d5bb13750640", "messageHeadline": "Move ort flatbuffers header to use enum class instead of enum (#5105)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 149, "deletions": 148}}, {"node": {"oid": "433061531e5fcd72e4174b15e1ce22469c8e948e", "messageHeadline": "Enable onnx_test_runner for ort format (#5100)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 109, "deletions": 16}}, {"node": {"oid": "62848c4de5ee7e690039b176c9bfe5a6264c397d", "messageHeadline": "Add store builds to nuget packaging (#5040)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 292, "deletions": 129}}, {"node": {"oid": "9ba56dcfed19ea6367c3243276799311f842f490", "messageHeadline": "Support Send and Recv for old NCCL versions (#5097)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 27, "deletions": 13}}, {"node": {"oid": "09a6ce6bc0acc19449aa72ec906a174ee2ad5e68", "messageHeadline": "Add re2 to memory leak checker whitelist (#5101)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "934f30fc38c680b0099ef44710e3997eb9b38963", "messageHeadline": "Not to call NVTX when not available (#5095)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 10, "deletions": 5}}, {"node": {"oid": "4b7aa16ed299e9c75f5d3a46029a1055af9c05da", "messageHeadline": "Fix a few more signed/unsigned warnings. (#5098)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "5e10cde0062c85b6eb6c7f368739a6c9f5a6f765", "messageHeadline": "PipelinesForCuda11Cudnn8 (#4938)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 401, "deletions": 9}}, {"node": {"oid": "a90fae8c712300152188fe3efab2480f86b94b52", "messageHeadline": "unify error handling in pipeline transformer (#5039)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 73, "deletions": 75}}, {"node": {"oid": "61151af321832f76649aa375a5d61feda2e1a210", "messageHeadline": "Fix typo in DML native method call from the C# API (#5083)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 28, "deletions": 15}}, {"node": {"oid": "f7c3e4fa99a87e1ea77fed4161d2df3152cce6f0", "messageHeadline": "Store/containerized apps support (#4651)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 75, "deletions": 28}}, {"node": {"oid": "924ecb062335d1b5f26d9ebf0a5667e73e1d5957", "messageHeadline": "Use manylinux2014 for Linux CPU build (#5091)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 14}}, {"node": {"oid": "6594d6672f40936933d72c14f560880c6085b32d", "messageHeadline": "Move onnxruntime.experiment to onnxruntime.training namespace (#5045)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 34, "deletions": 32}}, {"node": {"oid": "4ccca20def4c8a99b2afdb55ff109cbc5fc72b2e", "messageHeadline": "Replace MPI Send and Recv with NCCL Send and Recv (#5054)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 651, "deletions": 23}}, {"node": {"oid": "dbf4e7019d1b51212b00b12c7b0cc4587a819e1c", "messageHeadline": "Add ability to generate configuration file with required operators. (\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 110, "deletions": 47}}, {"node": {"oid": "80ada0291f9a8159787fead6108c196f6c34163d", "messageHeadline": "Improve the minimal build size on android and linux (#5086)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 30, "deletions": 15}}, {"node": {"oid": "5019b2f3b96fc5f3ca073003d483b6ebfb32535b", "messageHeadline": "fix for x86 android build break (#5088)", "author": {"name": "Guoyu Wang", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "a1a81470e36efdf5deacafaf2a014ebe5d5d6e83", "messageHeadline": "Add minimal build binary size verification (arm64) to Android CI (#5087)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 42, "deletions": 22}}, {"node": {"oid": "b8d63f31c3c742b5ee545d9ebfc9c874c775f574", "messageHeadline": "Bump bl from 4.0.2 to 4.0.3 in /nodejs", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "07bf8b968e31483d86d9bd5d88cd3b4bd903bb16", "messageHeadline": "Register BiasGelu and BiasDropout for CUDA only. (#5060)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "f41614a87532bd7b01a928a16ec9aacfcdd9b263", "messageHeadline": "User/brianma/telemetry (#5084)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 61, "deletions": 24}}, {"node": {"oid": "1b46573bb7a6fbb2c72b8e453c734a0b171f3fc5", "messageHeadline": "Update BUILD.md (#5085)", "author": {"name": "Moshe David", "email": "mosdav165@gmail.com", "user": {"login": "mosdav"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "a40d34386ad1a80648589acf08acd034c80dd7f2", "messageHeadline": "Add Linux CPU CI for ORT minimal build (#5074)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 69, "deletions": 0}}, {"node": {"oid": "b23e08b85c2792dbd8183bb8875f6ec5ff3d14f1", "messageHeadline": "Add AutoModel selector in transformers tool (#5051)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 59, "deletions": 18}}, {"node": {"oid": "4553b2eecd0a43de26a620d5d4b543c9f7b73b9a", "messageHeadline": "Expose DirectML provider to python (conflicts resolved from #3359) (#\u2026", "author": {"name": "Cameron Maske", "email": "cameronmaske@gmail.com", "user": {"login": "cameronmaske"}}, "additions": 37, "deletions": 3}}, {"node": {"oid": "c239ff0750fb2e626e3d5f3a154c99551d543cb9", "messageHeadline": "Modify embedlayernorm fusion due to shape node merging (#4967)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 182, "deletions": 34}}, {"node": {"oid": "38453acae3d90284e04813b8efc22f9cdada0fa8", "messageHeadline": "Further populate Stop Gradient list (#5021)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 28, "deletions": 2}}, {"node": {"oid": "e1ed0fde2bbcf2b56898a0ca456f2fc2981c4f64", "messageHeadline": "Prevent registering both DML and CUDA EPs in an ML op test (#5078)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 12, "deletions": 1}}, {"node": {"oid": "8d91d4ff3628634807ded024ba92abaf56cbceab", "messageHeadline": "Build docker image instruction fix (CUDA) (#5070)", "author": {"name": "Olivia Jain", "email": "oliviajain@berkeley.edu", "user": {"login": "oliviajain"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "6c33e95b886d3cabc9c0fa5a9139b97f4676b1c1", "messageHeadline": "Fix signed/unsigned mismatch on x86. (#5079)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "796ddeb2cba0f67ac723426906aed1f0163eb353", "messageHeadline": "Remove serialization of outer scope value info in ORT format model (#\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 45, "deletions": 48}}, {"node": {"oid": "e03a391895f08cb3fc5aefed6bd95d4cd840c986", "messageHeadline": "Small updates to ORT Mobile documentation (#5075)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 40, "deletions": 18}}, {"node": {"oid": "36dc057913f968566eaa1646cb5db41d8c5e7654", "messageHeadline": "Add unit test for C# setting of session options config entry. (#5073)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "5d60d57ce2b263f3e148c7b7d97b96c748405660", "messageHeadline": "Add csharp API for AddSessionConfigEntry (#5072)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "2c1410afe760af6ffded1419b097217713dc7f15", "messageHeadline": "Remove usage of macros for constants in public header. (#5061)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 21, "deletions": 23}}, {"node": {"oid": "6081c1cfa2d5db6d3d87fab1062210d6d78b5581", "messageHeadline": "Update ONNX to latest (#5069)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "d922cb1081b21eb05fa74c627cefb9f1d707f024", "messageHeadline": "Add sequence and map support in ORT mobile file format, add UT (#5066)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 461, "deletions": 142}}, {"node": {"oid": "de58720a972f5c6c9ad15c084b1d192de1ad3042", "messageHeadline": "Liqun/transformer test and e2e golden numbers (#5064)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 37, "deletions": 27}}, {"node": {"oid": "84de14a833b1e928c9dc74640197a4780abfe8cd", "messageHeadline": "Register OpSet13 CUDA Kernels for BERT/UniLMv2 (#4856)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 1212, "deletions": 422}}, {"node": {"oid": "370d194db7fe74e68edd0165ea42aa9c23204bc2", "messageHeadline": "Add a docker file for CI build CUDA 10.2 (#5065)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 35, "deletions": 0}}, {"node": {"oid": "ec88f14a7a4d5a058135ed63ab6d7e9df0d34e9c", "messageHeadline": "Implement QLinearMul in mlas (#4593)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1019, "deletions": 338}}, {"node": {"oid": "b5c2932ae88ec628dc8761dce91811dd1f22b033", "messageHeadline": "Last major set of ORT format model changes (#5056)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 983, "deletions": 223}}, {"node": {"oid": "6134994db93ccb5823be957e7218bde7df647f11", "messageHeadline": "Parallelizing elementwise kernels (#4577)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 354, "deletions": 180}}, {"node": {"oid": "0dad79b495cd9506fc18c5973114009546c5f9e2", "messageHeadline": "Add SetLanguageProjection C Api and use it in four projections (#5023)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 105, "deletions": 3}}, {"node": {"oid": "6dd4af3936c57cd0c46abf40eb150a068a05083e", "messageHeadline": "Fix initializer name only when wrapper is applied (#4920)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 49, "deletions": 55}}, {"node": {"oid": "d792af776d999851e3227d0ba705b1d3809a7bc8", "messageHeadline": "Remove Cuda dependency from TensorRT shared provider (#5014)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 751, "deletions": 389}}, {"node": {"oid": "78bb53381bb08c0c5d625c6e57dbb5c521dbd78d", "messageHeadline": "optimize resize op for NN mode for some fasterrcnn model (#4825)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 129, "deletions": 59}}, {"node": {"oid": "8289981f0e40f8cfaad661c35a2c094972e7ab2d", "messageHeadline": "Implement QLinearSigmoid. (#5015)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 201, "deletions": 32}}, {"node": {"oid": "0fc9c504fe687d7945d5b78494137199d07e7080", "messageHeadline": "Re-enable CI tests for the new PyTorch frontend (#5017)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 39, "deletions": 25}}, {"node": {"oid": "bd215b79a2aeeb5e0b77feaaed67cc7e3ac98ead", "messageHeadline": "ACL v20.02 (#4981)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 132, "deletions": 35}}, {"node": {"oid": "73456f10cd70554db7cd926dff9fd0fc83555fc9", "messageHeadline": "Fix contrib ops unregister to match pytorch behavior (#5052)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "d7502eff8f93b693d17d0654ba640e7a58153f8e", "messageHeadline": "Add nodejs samples README (#5005)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 15, "deletions": 15}}, {"node": {"oid": "4b9b5b6146dfae96799a05c44c8d4d05e3fc97ca", "messageHeadline": "Imported protoc cannot have compile options. (#5030)", "author": {"name": "xkszltl", "email": "xkszltl@users.noreply.github.com", "user": {"login": "xkszltl"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d7984fe6bad25212eea0fd7139af319796775d98", "messageHeadline": "Add packages from training docker to cgmanifest. (#5033)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 41, "deletions": 1}}, {"node": {"oid": "bb13b522911be49918790ba12f9420fbf4451cbd", "messageHeadline": "to allow parallel training with mpi4py (#4942)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 27, "deletions": 9}}, {"node": {"oid": "9388d49c0d66ed966abdfb970e8a72923df746fd", "messageHeadline": "Add warning to non pickable models (#5037)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 126, "deletions": 79}}, {"node": {"oid": "9d1bdef195e869ca75a6c52e114b7fce980c1be9", "messageHeadline": "Update CODEOWNERS and minor docstring fix (#5002)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "546965c2daf031a4b3bf08ad3125cdf5b9c941db", "messageHeadline": "Add deterministic path for AllReduceL2 (used to compute gradient norm\u2026", "author": {"name": "Suffian Khan", "email": "sukha@microsoft.com", "user": null}, "additions": 103, "deletions": 29}}, {"node": {"oid": "9ba2cfb71bb15e5d3a26aaacd940fb340bb97b86", "messageHeadline": "fix py packaging pipeline (#5038)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 39, "deletions": 0}}, {"node": {"oid": "22ba266bd6036c3d01659d51dcf248e08137666b", "messageHeadline": "Add flag to _internal_use to control export of contrib ops in ort tra\u2026", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 63, "deletions": 3}}, {"node": {"oid": "28445c88f926aa255de740836b00d741eb7a7263", "messageHeadline": "Changes to enable saving and loading an ORT format model (#4995)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 864, "deletions": 383}}, {"node": {"oid": "bbb9d92a5fe12a36d5fb6f9b7f53548b77eaa2e8", "messageHeadline": "Remove SchedulingParams variants of ThreadPool::TryParallelFor (#5050)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 21, "deletions": 117}}, {"node": {"oid": "fde7a2c84800309d34eeb2f77d6c8f27014f477f", "messageHeadline": "Temporarily switch SafeInt to a fork for an option to disable excepti\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "e0d1cf19a6e54867dc5f15a2977d3685fbdb417c", "messageHeadline": "Fix allocator bug (#5042)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "32687176151128fe590a2250f9948adc4639cb76", "messageHeadline": "Enable TF32 for training on A100 (#4914)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 45, "deletions": 8}}, {"node": {"oid": "a9db287bd70b1bfd49ca7ea26e45ebedda13a6c5", "messageHeadline": "Return windows error code for library loading and unloading failure (\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 17, "deletions": 16}}, {"node": {"oid": "b4e9e98cee1c7695aae6bb695d18257ef0112774", "messageHeadline": "Add more huggingface models in benchmark tools (#4986)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 138, "deletions": 23}}, {"node": {"oid": "a935731bd3e01fcd753b1224b70ed9c8b4c2eb49", "messageHeadline": "Neg Gradient (#5022)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "4a0f6595eb38f4f9fca7c14548d999876fbc1282", "messageHeadline": "Enable metadata and signature changes in graph transformers (#4783)", "author": {"name": "Dudeldu", "email": "mustermann.informatik@gmail.com", "user": {"login": "Dudeldu"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "4fd4b7414968fd1c0ca486cb63897236106f3fe3", "messageHeadline": "Change session option values if they don't work with EPs being regist\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 31, "deletions": 24}}, {"node": {"oid": "8a03b6e5c7adebe9bf971d4d3762df7e8c08a7f2", "messageHeadline": "Render Operator documentation as compliant markdown (#3658)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 607, "deletions": 939}}, {"node": {"oid": "e1901a7e10b4a156093288d0c0a200700f3caa5e", "messageHeadline": "Improve performance of CUDA implementations for GatherElements and Gr\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 329, "deletions": 241}}, {"node": {"oid": "d5d5e37e76cd3ed72925601861871322f6745971", "messageHeadline": "Build system enhancements (#5012)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9555, "deletions": 55}}, {"node": {"oid": "aabed34d5cd8d06ab80f496aec9e9a912e979e15", "messageHeadline": "Fix checkpoint API and improve loss scaler handling (#4950)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 256, "deletions": 113}}, {"node": {"oid": "eebc2cccce62e5893053fcf061f832168c6b2e55", "messageHeadline": "Fix fetches when eval_step's input is a subset of train_step's input \u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 145, "deletions": 2}}, {"node": {"oid": "a6e219deff9ad90fa4afc664af2cb5f90a93b6bc", "messageHeadline": "Pass Model Path to TensorProtoToMLValue from Constant Folding for Ext\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 12, "deletions": 7}}, {"node": {"oid": "5651c232718cb4cb2865cc32729ace4bfdecd76f", "messageHeadline": "Fix for Android ORT android initOsArch exception (#5006)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "44b3accb74de5bcff07189704724b7f3012ec8f9", "messageHeadline": "Missing header for `std::once_flag` and `std::call_once`. (#5010)", "author": {"name": "xkszltl", "email": "xkszltl@users.noreply.github.com", "user": {"login": "xkszltl"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "9902b57090fae6f30c83ea49c14f75137ab9a1a1", "messageHeadline": "Fix a warning in global_thread_pools/test_inference.cc (#4987)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "f38f2d5b5495acdf8fd21c25976d67df47a3f1bf", "messageHeadline": "Port #4920 into the new pytorch frontend (#4965)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 42, "deletions": 53}}, {"node": {"oid": "d30dd41c0e26d85c98dcd4b21a07789c8fe82cd3", "messageHeadline": "Remove public default ctor in PyInferenceSession and replace it with \u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "c6a3620ba8ebfd1e199eeb01f33d1419df40b337", "messageHeadline": "Remove evaluate telemetry due to redundancy (#4996)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 8, "deletions": 26}}, {"node": {"oid": "a47cae031f52a29086c8756dc2d255bb9912914d", "messageHeadline": "Use raw attention mask in BERT related fusions (#4889)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 93, "deletions": 178}}, {"node": {"oid": "d79af260bb5894c2a85efbf4b590e9b098f16286", "messageHeadline": "Liqun/new api orttraining test transformers (#4982)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 170, "deletions": 64}}, {"node": {"oid": "64237d999c6922a610df1fe604e904a5ad4e0e5f", "messageHeadline": "Add Cmake config for onnxruntime_NO_EXCEPTIONS (#4975)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 64, "deletions": 29}}, {"node": {"oid": "ad1701dfb1bdbcaefa56396863bc92b3f1859140", "messageHeadline": "Rename DeviceAllocatorRegistrationInfo to a more generic name; Use Or\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 243, "deletions": 245}}, {"node": {"oid": "ffc2b25a3ab9555dc1255d13c3a4e0b8e06191ca", "messageHeadline": "Quantization tool improvement (#4933)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2388, "deletions": 1703}}, {"node": {"oid": "464bbd27a939ebc73bfd7fe3eea0eeb93a76e56b", "messageHeadline": "Zhalei/optimize nms (#4875)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 56, "deletions": 22}}, {"node": {"oid": "cf1b74396a1133c8bbe2004358031a6f8ed5ba18", "messageHeadline": "Fix build break for microbench. (#4960)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "14b51d6502a743e1b6c779e55efb9f2c7368a950", "messageHeadline": "CiPipeline@ReducedOpsBuild (#4917)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 155, "deletions": 23}}, {"node": {"oid": "7ca8388dc98cbc23c7c165064bca6b3142d28d79", "messageHeadline": "[ORT Mobile] file format schema and file I/O code (#4973)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 3610, "deletions": 109}}, {"node": {"oid": "bca9ccb1b387893bbfe3ad9c78bee2eb42e51b55", "messageHeadline": "add install sec updates (#4957)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "1e1f5a9c7917374abd146dd4368a71c6111d4388", "messageHeadline": "support data parallel + pipeline parallel (#4648)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 869, "deletions": 122}}, {"node": {"oid": "9817b8c8a76c1e08d6d7597932d81bd77d81638f", "messageHeadline": "Fix state_dict/checkpoint issue introduced by #4639 (#4984)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 9, "deletions": 10}}, {"node": {"oid": "8679a7244e92d9af1d866295d53320871e953794", "messageHeadline": "Enable rejecting models based on onnx opset (#4912)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 222, "deletions": 29}}, {"node": {"oid": "50c610e70a2c1fc3e304199d1fec240dd8c029f1", "messageHeadline": "Stop Gradient at Shape op (#4983)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "7af052fd62a595039dfac49164b984a13c58d008", "messageHeadline": "Add CI status badges for Training builds (#4951)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "6d9d252bc3aa9821829c3dfccac5c6e33e3b363d", "messageHeadline": "Disable NegativeLogLikelihoodLoss_LargeSizeTensor test (#4979)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b41e5e88fb649799c28191922feaaf5e1428e951", "messageHeadline": "Add more node debug dump functionality. (#4921)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 647, "deletions": 163}}, {"node": {"oid": "98f7fdd7da7de7ab513d6ce851f824e5f2742d5e", "messageHeadline": "Handle MatmulGradient with 2D Weight at B (#4977)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 51, "deletions": 24}}, {"node": {"oid": "bac41969be76d2e333df98e321c38b65b301095b", "messageHeadline": "update (#4948)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "64d52ae47d1d8a1d8b3e768ccd7e818f6503354a", "messageHeadline": "Support creating sessions using DML EP via C# (#4955)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 20, "deletions": 3}}, {"node": {"oid": "7080e485a334ec58ecbdac7a691632049bcc3e90", "messageHeadline": "hHandle upper-cased subscript labels in Einsum (#4964)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 39, "deletions": 6}}, {"node": {"oid": "f4b057b0986cf1e3d0ea57cbdf482435c18e4764", "messageHeadline": "Fix DML License in nuget package (#4969)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ea5732319e3bfd7e81b309fbd72c78ca62f86fd3", "messageHeadline": "Add option ORT_NO_EXCEPTIONS to disable most exception/throw in /onnx\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 990, "deletions": 573}}, {"node": {"oid": "655ffd5d5b6b17d3acd5c759fa90d4ae8924ecdf", "messageHeadline": "make (de)tensorization events measure level events (#4958)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 143, "deletions": 14}}, {"node": {"oid": "cd0f2fb48c0c864640b14e5b26600422b39fbb23", "messageHeadline": "Add code oweners for pytorch frontend (#4963)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "7045910d1057a20d54de82fea31bf45d12818f15", "messageHeadline": "Support RegisterCustomOpsLibrary via the Python API (#4764)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 349, "deletions": 125}}, {"node": {"oid": "040c5fa3e04d47b337fc99c58dfac9f785f003ac", "messageHeadline": "Merge pull request #4925 from microsoft/user/dwayner/Iron", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 2954, "deletions": 875}}, {"node": {"oid": "1281ff64625939c445331a458793e4e64fd5be27", "messageHeadline": "Put operators in-between Wait and Record (#4916)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 814, "deletions": 1136}}, {"node": {"oid": "b945225de3f2830565c89719d911cc609c4ddb4c", "messageHeadline": "Include DirectML pdb in x86 bin folder (#4953)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "c37fa7c2781f0a1d7d38fbc3b4609bef05834844", "messageHeadline": "Delete Dockerfile.centos6_gpu (#4851)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 280, "deletions": 180}}, {"node": {"oid": "39382dc6c3456a0dbeb4f68f9d41760ac4abc791", "messageHeadline": "Update winrt_api.md to address the 1.4 release (#4946)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "79429c934b76cefb475fa507be7e82f95c7d5039", "messageHeadline": "Update", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "a7ce5b2be183fbac33904b609ee2f604a80dc2e7", "messageHeadline": "fix comment and casing of telemetry fields for named dimension overri\u2026", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "dfb9d97ddfed58448d833df660752a24b41c71a9", "messageHeadline": "Support DistilBert's Attention fusion in Optimizer (#4748)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 812, "deletions": 280}}, {"node": {"oid": "e6b6736e481f4f233afb523805e9b455854715ce", "messageHeadline": "update cuda capabilities (#4936)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "efdd96595f5120af20ac8ffb8b1ad0917e8553da", "messageHeadline": "bfloat16 and opset13 related fix (#4913)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 101, "deletions": 41}}, {"node": {"oid": "f68d5263b781d047ba85722ea33856d566e05232", "messageHeadline": "Merged PR 5100436: EinSum ONNX 1.7 (opset 12) ORT DML EP kernel", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 531, "deletions": 40}}, {"node": {"oid": "b5c765c76beda51b876c162645a3ef6e23712fe6", "messageHeadline": "Merged PR 5103319: 8d Update", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 45, "deletions": 37}}, {"node": {"oid": "970ddd56a7c411324b1257691eb57ad913fbcf70", "messageHeadline": "Fix typo in contributing.md (#4939)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9f5d4918dc3dc8d2e69ac2e8c369993cdcd4aac3", "messageHeadline": "MatMul Gradient optimization for dB when B's is 2D tensor (#4899)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 78, "deletions": 77}}, {"node": {"oid": "6dc85b5f14f709458da687521394e1265898eb1f", "messageHeadline": "wstring_convert std::codecvt_utf8 add ~200KB to inbox windows.ai.mach\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2b460eaecad1dc3faabb3bc65aa9553670052637", "messageHeadline": "Revise IDisposable implementation in C# interfaces (#4915)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 210, "deletions": 257}}, {"node": {"oid": "08eb15068cf5298643097e144591f88845974b9d", "messageHeadline": "Exclude the Map types from the build if ML ops are disabled. (#4908)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 162, "deletions": 25}}, {"node": {"oid": "792ed4453721d8d2b5138e8598ebc8dda36f73cf", "messageHeadline": "Support EmbedLayerNorm fusion for DistilBert (#4928)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 475, "deletions": 228}}, {"node": {"oid": "00fe71826428307fccb832f45bb633e342e8d332", "messageHeadline": "Fix divide-by-zero for SSCE kernel when normalize factor is zero. (#4\u2026", "author": {"name": "harshithapv", "email": "54084812+harshithapv@users.noreply.github.com", "user": {"login": "harshithapv"}}, "additions": 20, "deletions": 4}}, {"node": {"oid": "cac25751bdd95dadb3f81d196421539914731f8c", "messageHeadline": "Fix mnist example (#4926)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "438babd966278ae1331bffbf369429199b0bd028", "messageHeadline": "Fix some Android build issues when ORT_MINIMAL_BUILD is defined. (#4924)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 13, "deletions": 17}}, {"node": {"oid": "b3783a9f85ba1ac6b3d1f45a18f4727791382001", "messageHeadline": "matching multiple choice between new and old apis (#4918)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 90, "deletions": 36}}, {"node": {"oid": "0d3bbfdd0f0ddb38eea799a52fcd2dd66513e5b7", "messageHeadline": "enable nuget packaging in local builds (#4884)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 395, "deletions": 74}}, {"node": {"oid": "0a2848d3a012d8110dfc8274455632bbb3508d13", "messageHeadline": "Remove cerberus from wheel package (#4919)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "cb5e199a790239b200cbb601f1493a602d6e0ba8", "messageHeadline": "Merged PR 5093868: GatherND1 ORT DML EP", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 119, "deletions": 12}}, {"node": {"oid": "5d3638e93534fd313212098d04788900e6913901", "messageHeadline": "Fix symbolic shape inference bug when subgraph contains Constant node\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 2, "deletions": 0}}, {"node": {"oid": "170fee0987ecd50879a774158e6898ca1fc688b9", "messageHeadline": "User/xianz/fixbuild (#4906)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "1161c4d75feee979ce054111cf2110b6916a8bcf", "messageHeadline": "Exclude MLAS AVX512 in minimal build (#4905)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "cb2dfee31c3cf5867aca936e887e489e5699fa34", "messageHeadline": "Size Op - CUDA kernel support (#4868)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 25, "deletions": 0}}, {"node": {"oid": "294eaca9eff6f6d750a03fbea3451b155d360029", "messageHeadline": "Support double for ArgMax operator (#4907)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 56, "deletions": 28}}, {"node": {"oid": "3d63d8d4f141433135ecbf19e7705b00e6218954", "messageHeadline": "Extend C++ API for Map/Sequence Type Info (#3517) (#4781)", "author": {"name": "Dudeldu", "email": "mustermann.informatik@gmail.com", "user": {"login": "Dudeldu"}}, "additions": 114, "deletions": 0}}, {"node": {"oid": "6c26e521346a246cdcd109412b0bc41a49cfafd0", "messageHeadline": "Support accessing a model's metadata in C# (#4867)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 307, "deletions": 13}}, {"node": {"oid": "26bd8c20854aafbc36d843892eb2eac2440b6065", "messageHeadline": "Support scalar tensors in c# (#4849)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 112, "deletions": 26}}, {"node": {"oid": "d3cddba8f16a820eded2914deaa78b6bd5780f5b", "messageHeadline": "Add this line to allow collection of AppSessionGuids (#4901)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 1, "deletions": 0}}, {"node": {"oid": "14c691030f2f0a8de896463e29a8dd342256fc11", "messageHeadline": "Fix build break from removing custom ORT onnx protobuf (#4904)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 19, "deletions": 3}}, {"node": {"oid": "71d8846635a20a97d85a27e4fffcbad05ed00813", "messageHeadline": "Fix telemetry-steps.yml (#4903)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 11, "deletions": 20}}, {"node": {"oid": "f34ed3a57640506ec357c7577b8b2274de7f53cc", "messageHeadline": "Hot fix for the python packaging pipeline Linux ARM build (#4902)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "83b7c1151a5840ccbe33568a9ae9b0b43a9b9fcf", "messageHeadline": "Merged PR 5086432: Update Operator Pow-12: Allow exponent to differ f\u2026", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "db6a821869abdd4149f5f0635e7d183d913d143c", "messageHeadline": "Enable example transformer test with dynamic size inputs (#4888)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 40, "deletions": 6}}, {"node": {"oid": "824fcbfd9dda0e07f8e0f4f4529a4b2f6bd4866e", "messageHeadline": "support Normalized_0_1 and Normalized_1_1 (#4800)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 317, "deletions": 95}}, {"node": {"oid": "268d2283c04e6746ca7d9018273c02efb25e4a0c", "messageHeadline": "Export GPT-2 ONNX model without postion_ids and attention_mask inputs\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 227, "deletions": 130}}, {"node": {"oid": "26546f81fe8b99129b1420181ce86b0f3c386283", "messageHeadline": "Remove the private ONNX protobuf definition file (#4878)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 85, "deletions": 288}}, {"node": {"oid": "c5cb9d7b411cbf647d3046b62a43969af894d545", "messageHeadline": "match reshape fusion for distilbert (#4844)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 227, "deletions": 25}}, {"node": {"oid": "744809ceaed2cda2b4874fdd67790c819f18a8b9", "messageHeadline": "Detect whether the node has been inserted cast nodes twice (#4811)", "author": {"name": "Chun-Wei Chen", "email": "jacky82226@gmail.com", "user": {"login": "jcwchen"}}, "additions": 43, "deletions": 29}}, {"node": {"oid": "1542a8a9b78dd42f1921d28f95a949118a4f025c", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 12, "deletions": 1}}, {"node": {"oid": "3dd2d53593276a153b69c3e1d2a287b80171282d", "messageHeadline": "Merge branch 'DmlDev' of https://microsoft.visualstudio.com/DefaultCo\u2026", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "47c4144bd1178daf31006726eef2b4a85eee15c6", "messageHeadline": "Add gcc/clang flags to make binary smaller (https://interrupt.memfaul\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 1}}, {"node": {"oid": "8cc5e9748f43a5fb2a12d4b788d251520dc374bf", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1055, "deletions": 320}}, {"node": {"oid": "7c0ff3742fe5fd205d2906a9c5ec000be5d5de4b", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1055, "deletions": 320}}, {"node": {"oid": "eb05db5a2ab989a7696c51beeb566e2f13a8b85e", "messageHeadline": "Fix OptimizerConfig params groups (#4877)", "author": {"name": "Rayan-Krishnan", "email": "rayankrishnan@live.com", "user": {"login": "rayankrish"}}, "additions": 156, "deletions": 43}}, {"node": {"oid": "728e886bba0e5dbf264bcf921b2ec86b19fea981", "messageHeadline": "Add kernel def hash logic for minimal build (#4891)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 162, "deletions": 62}}, {"node": {"oid": "db7669b22576309a7d954a1232dff73c503d8e3f", "messageHeadline": "Reduce ONNX dependency in minimal build (#4890)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 169, "deletions": 108}}, {"node": {"oid": "29dcfb24ab13ffa585131acd1f623f7ca5da073b", "messageHeadline": "Allow multiple sessions to share an allocator, optimize constant fold\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 575, "deletions": 114}}, {"node": {"oid": "46863a2868db3477b3c7b72e03bf20f84ec1117c", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2529, "deletions": 991}}, {"node": {"oid": "fa68bbc82e030c8f4cacba8a222bbedbc2107042", "messageHeadline": "Relu grad kernel (#4864)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 119, "deletions": 62}}, {"node": {"oid": "dce2ce7a4f6e865045a40df8760e84239ca2ec92", "messageHeadline": "Fix checkpoint API and copy samples into build dir (#4887)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 33, "deletions": 34}}, {"node": {"oid": "6260d073b3ba274086fa67ffefd0d87112bac4d2", "messageHeadline": "Glue parallel training (#4550)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 281, "deletions": 150}}, {"node": {"oid": "9a6db9b9f4b019669675a7b145e5e2a871be4eef", "messageHeadline": "Fix next node access bug in calibration tool (#4863)", "author": {"name": "RRRachelllll555", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 55, "deletions": 10}}, {"node": {"oid": "3fa73a5b6ace4cad28cd6b6a6e0e795acc08371a", "messageHeadline": "ReduceBinarySize (#4747)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 455, "deletions": 14}}, {"node": {"oid": "82bc21e35e7aed6ee93d06381096f7e71c2a3786", "messageHeadline": "Namespace change on ort flatbuffers schema (#4886)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 42, "deletions": 44}}, {"node": {"oid": "fdd0926d0050b85c822f2231dd1d260ddd5449a8", "messageHeadline": "int64_t support for GatherND cuda (#4881)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 19, "deletions": 2}}, {"node": {"oid": "acbf6d15c6e1d917fffaef5a58e9c772b124135a", "messageHeadline": "Improve LRScheduler tests (#4885)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 175, "deletions": 102}}, {"node": {"oid": "e00ad83f2b3181eb6e3901fcf78a0d00c0306a04", "messageHeadline": "Initial changes to disable code in a minimal build (#4872)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 588, "deletions": 343}}, {"node": {"oid": "fb43aa0de07604860f7b4c7586fc9bd506a188b6", "messageHeadline": "implement per-channel for quantizelinear and dequantizelinear (#4759)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 620, "deletions": 114}}, {"node": {"oid": "5427a7e9af5dae4b3ba016511051bdd7f3c56c74", "messageHeadline": "Update LRScheduler to use scheduling similar to HuggingFace (#4880)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 181, "deletions": 155}}, {"node": {"oid": "985b741780263641310e7ca1361877f994bfcf5c", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 4214, "deletions": 1969}}, {"node": {"oid": "ef19916d0781c0d50c6e4a0deca0f0bbe75269f3", "messageHeadline": "Add Node::SinceVersion() (#4874)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 31, "deletions": 20}}, {"node": {"oid": "d4d52056be03fc4aa63039ca305d4b3cef24d581", "messageHeadline": "Flatbuffers schema for serialization of the onnxruntime::model/graph \u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 222, "deletions": 3}}, {"node": {"oid": "a0271f619a29df7a8ec59ae6a4c224b66f01cfe8", "messageHeadline": "Range CUDA inputs should be in CPU (#4871)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "c5342b5417ea5a4baeb3f5b1e7ef0b7f68e9b9d0", "messageHeadline": "fixed compilation issue for Jetson Xavier (#4873)", "author": {"name": "paradigm", "email": "RubenProkscha@gmail.com", "user": {"login": "paradigmn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "730d95107dc42f296246146e1911bfa2fc794551", "messageHeadline": "Merged PR 5065263: ONNX backend autopilot crash in WinML ToString run\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 12, "deletions": 9}}, {"node": {"oid": "40324e6f779e238dda18f3d7de7149ef1f461ec1", "messageHeadline": "Merged PR 5068977: RoiAlign (ORT DML EP)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 876, "deletions": 117}}, {"node": {"oid": "c6119a548ce42c7c5bd020e2b2efc70df7ce431a", "messageHeadline": "enable telemetry in node.js binding", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3a00b50cf8cdb1662d3921366901343b6cc5fc2c", "messageHeadline": "[OpenVINO-EP] Updating OpenVINO EP to 2020.4 (#4836)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 2542, "deletions": 1205}}, {"node": {"oid": "d00a70a4320292342b22638c961d9bf8bd63dc40", "messageHeadline": "Fix broken Nuphar docker file by removing stale build options", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7589445e6e454cc202f6a33cf79d7da223b2fb21", "messageHeadline": "Add ONNX BERT Frozen Weights and Save as ONNX Tests (#4859)", "author": {"name": "Rayan-Krishnan", "email": "rayankrishnan@live.com", "user": {"login": "rayankrish"}}, "additions": 100, "deletions": 3}}, {"node": {"oid": "25cc6158a8a57f507511524d1b3bc210f65808bc", "messageHeadline": "update golden numbers (#4865)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 16, "deletions": 16}}, {"node": {"oid": "101cd80bd7fba5d8ac1a4b5d68f9b0c73a7a869e", "messageHeadline": "fixed a warning on an unreferenced local variable (#4861)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dc50aa42d5b1cf591b8d63ec37bdf108c389da39", "messageHeadline": "Refactor session state finalization and kernel lookup usage  (#4763)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 459, "deletions": 470}}, {"node": {"oid": "d7233c7c9756e572ce42446750a37d1dbc01654d", "messageHeadline": "Fix training for models with dict input (#4842)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 157, "deletions": 100}}, {"node": {"oid": "7cc88ef7ed6ba0cb48688c2d17592e627ff456fa", "messageHeadline": "Port legacy checkpoint API into new front-end (#4855)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 594, "deletions": 104}}, {"node": {"oid": "c8155c79cbd31bc4c65426b5d963bf6d0d34912c", "messageHeadline": "Merged PR 5051981: ArgMin/Argmax 12: add select_last_index attribute", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 165, "deletions": 16}}, {"node": {"oid": "75ad7be3369e57314e12d6ce7ecd7940252a9f75", "messageHeadline": "add caching support for dynamic input models (#4702)", "author": {"name": "Alexandros Koumparoulis", "email": "66834395+alkoumpa@users.noreply.github.com", "user": {"login": "alkoumpa"}}, "additions": 77, "deletions": 32}}, {"node": {"oid": "d350d9ac5ef2d83864e68574aa01bcb7b0b37ebe", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 635, "deletions": 514}}, {"node": {"oid": "fff0b41fcb7cb85321f65d68c1dedaaf7032fcb0", "messageHeadline": "Nuget build break fix (#4854)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "5eaac31faa291ec994963c7accbd8f2c4ca5ca5a", "messageHeadline": "support opset13 on transformers. (#4837)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 137, "deletions": 126}}, {"node": {"oid": "61a5502af0c4ce36dad1a9f5616c381600ccfaf9", "messageHeadline": "Fix some incorrect operator registrations. (#4838)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 182, "deletions": 142}}, {"node": {"oid": "1ba07ccfaf35da2f1cde696a20c9ae87e470efe9", "messageHeadline": "Codesign validator fixes", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "05758819497aa1d90fc197cfa6315ffbdcc0bd81", "messageHeadline": "Update quantization notebook to pytorch 1.6 (#4834)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 81, "deletions": 73}}, {"node": {"oid": "dee7596724bac5232bf6cfff66c6018439a12576", "messageHeadline": "Add a generic collection of session configurations to the SessionOpti\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 201, "deletions": 147}}, {"node": {"oid": "81ff1688339db26771f008bfd96e0955b135594f", "messageHeadline": "Update stale.yml with current labels and mark stale items as \"stale\" \u2026", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "2605af9a0bbc64e315177b6ab6e52d5ec1c2cee5", "messageHeadline": "Fix for mainz model (#4744)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "f3b0c93a455cf974e73a9c82bb0caa75427a3c81", "messageHeadline": "Fix issue preventing loss scaler to run due (#4833)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 20, "deletions": 20}}, {"node": {"oid": "c4353b65cf9e0800a784aabfbf4b79363af25e3b", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 7249, "deletions": 695}}, {"node": {"oid": "a3c95374c33463b85db84e388488baf52bd2d9df", "messageHeadline": "Support asymmetric paddings in CUDA Conv kernel  (#4627)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 483, "deletions": 188}}, {"node": {"oid": "c878ecbbe09a5a9e9f52bb855c2378c23ef07aad", "messageHeadline": "Sahar/csharp support openvino (refined) (#4835)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 55, "deletions": 3}}, {"node": {"oid": "24d9f4e0c3edff7c2355492d21e2f368a49637c3", "messageHeadline": "Add More Extensive ONNX BERT Tests (#4827)", "author": {"name": "Rayan-Krishnan", "email": "rayankrishnan@live.com", "user": {"login": "rayankrish"}}, "additions": 596, "deletions": 28}}, {"node": {"oid": "e98697ec28f4b74754532e905f08a39307264aef", "messageHeadline": "Fix nuget cpu package pipeline (#4832)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d3af6699802c818d5660aeaa4f662738af3ec7b3", "messageHeadline": "Auto upgrade base image dependencies (#4797)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "f933910ea3cb28cf322f5a467ec5d9e9ec9dbf95", "messageHeadline": "Update LambConfig defaults to match backend (#4826)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "6a360bad6bfb3dc1efe2d5095c58420689d5e9e7", "messageHeadline": "ReplaceStrncpy (#4823)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "32a5f3d5b62230153ef12b761a3e560c0133caac", "messageHeadline": "Check status of element-wise op prepare functions. (#4830)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "ef20efe015dfd725445e801f389c429706a1c99b", "messageHeadline": "Register cerberus license into ThirdPartyNotices.txt (#4828)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 23, "deletions": 0}}, {"node": {"oid": "ea37a4d89bf0f9a7c1d4884bd048d49cf4879c95", "messageHeadline": "Add Trilu custom op (#4537)", "author": {"name": "Ksenija Stanojevic", "email": "KsenijaS@users.noreply.github.com", "user": {"login": "KsenijaS"}}, "additions": 1137, "deletions": 37}}, {"node": {"oid": "1ce2982f65e5516067fdcaef19409279173b0d75", "messageHeadline": "Update GPT-2 notebook using IO Binding example (#4799)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 628, "deletions": 235}}, {"node": {"oid": "360e2ae11b61f71fd8544c5e5ea34427a2937988", "messageHeadline": "Update eigen to the latest to support C++20 (#4817)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "94a6f50af6eb25ea2379e4725f53c802f7562f06", "messageHeadline": "Revert \"Sahar/csharp support openvino (#4703)\"", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 54}}, {"node": {"oid": "42408aa3ed16df867f7d3b4fde0d8975a6c6e50b", "messageHeadline": "Add new PytTrch front-end (#4815)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 3740, "deletions": 4}}, {"node": {"oid": "5eec4f66edbfd6441056eed837432a42e191dcd2", "messageHeadline": "Refactor manylinux docker image and the related pipelines (#4751)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 643, "deletions": 213}}, {"node": {"oid": "bf300e3c189fcbf91562ef27f4a75fb2349744fa", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 471, "deletions": 156}}, {"node": {"oid": "ea3b4e1f8da7949fc90a6dbe07c6759ec4369a9b", "messageHeadline": "Fix bug in DispatchOnTensorType macro (#4808)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5899c1197ae19e9b01e97e937336a46cb63cd2fa", "messageHeadline": "add telemetry for named dimension overrides (#4794)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 22, "deletions": 0}}, {"node": {"oid": "0a0ac70eec67cecd11ace0b177bff687eff2a201", "messageHeadline": "Sahar/csharp support openvino (#4703)", "author": {"name": "sfatimar", "email": "64512376+sfatimar@users.noreply.github.com", "user": {"login": "sfatimar"}}, "additions": 54, "deletions": 3}}, {"node": {"oid": "1b1a6a4ca96b6a182028df0028504ff0ec2591a4", "messageHeadline": "Bump onnx to get bfloat16 in ops, and some update in ort to support b\u2026", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 394, "deletions": 152}}, {"node": {"oid": "6a638ad22ac34eadd5847146234f4b2bf289ad0b", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 11, "deletions": 1}}, {"node": {"oid": "8ba6b6a21e2b4c5fb3c440e9f5540643b1813091", "messageHeadline": "Support usage of C API with C++ standards older than C++11 (#4257)", "author": {"name": "Bogdan Bugaev", "email": "bbugaev@users.noreply.github.com", "user": {"login": "bbugaev"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "19494aba7a44fea21d5d9e8c5ab8e466cd371077", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 10, "deletions": 8}}, {"node": {"oid": "8d2e22558d6cf75530e581295daee579f20b9f89", "messageHeadline": "unattended-upgrades (#4804)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "5a8962d3270e77c55b619de5b66d7ea8e94a1d98", "messageHeadline": "Make grad name unique (#4788)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "56a5c7fc85fe8d87506405c562a205f93421cc7a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2135, "deletions": 83}}, {"node": {"oid": "afa89566d73f624189c24e6fd74deb32def27c3f", "messageHeadline": "Using cublasGemmBatchedEx/cublasGemmStridedBatchedEx for training (#4\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 239, "deletions": 52}}, {"node": {"oid": "ec36c793e84fb4465ee875c186a9989dc4d78774", "messageHeadline": "Eliminate redundant subexpressions (#3047)", "author": {"name": "Maxim Kalinin", "email": "makalini@microsoft.com", "user": {"login": "MaximKalininMS"}}, "additions": 982, "deletions": 8}}, {"node": {"oid": "ce65275edf8540248ba2b721ae2c7351c3284edd", "messageHeadline": "C# samples: Faster R-CNN (#4733)", "author": {"name": "Marcus Turewicz", "email": "24448509+marcusturewicz@users.noreply.github.com", "user": {"login": "marcusturewicz"}}, "additions": 408, "deletions": 0}}, {"node": {"oid": "de2685261bc83fa3b7871a75f12b93029754213c", "messageHeadline": "Install AzureML support and commonly used packages in the training im\u2026", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 455, "deletions": 2}}, {"node": {"oid": "7acef875bb9b1e5c2f92da96734ff593e2c537b8", "messageHeadline": "Fix bugs in TensorRT (#4780)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "aa993e95c96c839f483af14de83371df77dbd21a", "messageHeadline": "enable build flag '--use_openmp' on MacOS (#4774)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 20, "deletions": 9}}, {"node": {"oid": "f12e9de11124915f804432d3586efba32c13afce", "messageHeadline": "build fixes for https://github.com/microsoft/onnxruntime/pull/4721 (#\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 23, "deletions": 4}}, {"node": {"oid": "f01579b8fb60c5fbe0012e2fed12871801979339", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 13633, "deletions": 5363}}, {"node": {"oid": "aca34352a54794754bfa7532a9d577b67935cf9e", "messageHeadline": "Java API: Documentation cleanup (#4395)", "author": {"name": "James Yuzawa", "email": "jyuzawa@triplelift.com", "user": {"login": "yuzawa-san"}}, "additions": 32, "deletions": 33}}, {"node": {"oid": "722602f32d6f8d335f5a1dc73c9e4629ffc0a4a8", "messageHeadline": "replace namespace reference with alias (#4786)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5e7a6e78e37bbbbe45539789d0a83313386d7739", "messageHeadline": "Changes for BART dynamic shapes in reduction (#4730)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 70, "deletions": 57}}, {"node": {"oid": "74b3b8448c893318054aae4e31464ce00e040496", "messageHeadline": "Fix MatmulTransposeFusion::ApplyImpl() setting of modified flag (#4775)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 59, "deletions": 4}}, {"node": {"oid": "8fb743f767a3b1de867fc902cc5995833b3607ed", "messageHeadline": "Refactor Cast to reduce binary size. (#4765)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 233, "deletions": 325}}, {"node": {"oid": "9cec98ec1bcb6b8a668733fcd00d81af61b44d93", "messageHeadline": "Honor allow_spinning at barrier at end of parallel sections (#4767)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 93, "deletions": 8}}, {"node": {"oid": "61b2a663a3892f92ddd93af4f3e292d2f84da304", "messageHeadline": "Update Python version support (#4778)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cddddc4d5597e5d59dbc7d669c4800f2d1991ef7", "messageHeadline": "Add missing header file to MNIST.cpp (#4773)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "a69ca6389505a201a7de1693ee5128b00a021f24", "messageHeadline": "add --no_attention_mask option (#4750)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 23, "deletions": 2}}, {"node": {"oid": "adda8c66d997aedaeaff7c6f9d80eed0b6edaa92", "messageHeadline": "Docker image release pipeline (#4682)", "author": {"name": "jingyanwangms", "email": "47403504+jingyanwangms@users.noreply.github.com", "user": {"login": "jingyanwangms"}}, "additions": 95, "deletions": 12}}, {"node": {"oid": "8a66ad79a6c401eee4853776a7b37d6ff3919059", "messageHeadline": "Add Experimental WinRT API IDL as placeholder for adding new winrt fe\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 227, "deletions": 4}}, {"node": {"oid": "7e955960f1edfdad1fecd9967042348d281184ac", "messageHeadline": "Optimize Slice Kernel by Removing If-statement (#4753)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 70, "deletions": 50}}, {"node": {"oid": "b7254551f043bf8466d0d729dca05ecd67cb64d4", "messageHeadline": "Add new api function At() (#4457)", "author": {"name": "Josh Bradley", "email": "jgbradley1@gmail.com", "user": {"login": "jgbradley1"}}, "additions": 106, "deletions": 31}}, {"node": {"oid": "38c804a048d647025b967ac7c1837be5c2075a7a", "messageHeadline": "Fix broken link to ScoreMNIST.java in Java_API.md (#4213)", "author": {"name": "Scott Bonebrake", "email": "scottbonebrake@gmail.com", "user": {"login": "sbonebrake"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ac725b53f6405bc607a61860e983772c33472880", "messageHeadline": "Convert TensorRT provider into a shared library (#4721)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2025, "deletions": 1231}}, {"node": {"oid": "ac4997665ad0334e9e3f0996d4bd80ccaa925974", "messageHeadline": "Make Java Publishing and Java GPU pipelines to run nightly (#4749)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 17, "deletions": 1}}, {"node": {"oid": "f51385fd1e316ec74977e4273293554594facb34", "messageHeadline": "Yanchen/nuphar/clip 11 (#4737)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 26, "deletions": 7}}, {"node": {"oid": "6037b41c17e415aad99e0c880626cf3c4896918e", "messageHeadline": "Merged PR 5002755: Add Celu, GreaterOrEqual, LessOrEqual", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 147, "deletions": 3}}, {"node": {"oid": "3530ce541cbb66f05e523f92b62cebaa4793bd3f", "messageHeadline": "Expose IOBinding features via C/C++/C# language bindings. (#4646)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2967, "deletions": 1242}}, {"node": {"oid": "6c33d7f5df8006a8c7cb741ef114e2ff6c3feb41", "messageHeadline": "Fix bug in Loop optimization  (#4210)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 103, "deletions": 9}}, {"node": {"oid": "082a741636f689d7f9c0ff5c68fe9132072686d7", "messageHeadline": "Move DNNL workaround to EP (#4738)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 16, "deletions": 16}}, {"node": {"oid": "487665c21ff8d35417cfdf76a6d12f2240e4c728", "messageHeadline": "Transpose MatMul fusion fixes (#4728)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 181, "deletions": 22}}, {"node": {"oid": "316d1a9e69c0d961c94cbf19644c1ec910ccc5b9", "messageHeadline": "Update benchmark for large model or model name with non-alphanumeric.\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 143, "deletions": 74}}, {"node": {"oid": "6499a38b7d4a03dd9ed9db2a400676dbba2380ed", "messageHeadline": "Add the missing onnx_proto import (#4705)", "author": {"name": "Vagif", "email": "KyotoSunshine@users.noreply.github.com", "user": {"login": "KyotoSunshine"}}, "additions": 4, "deletions": 9}}, {"node": {"oid": "2e3ccc75187600c10ebdfa5cb62b538c987d1b00", "messageHeadline": "Change order of some checks to workaround a linker issue when /LTCG:i\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 10}}, {"node": {"oid": "24d4f7643674dc2ce614d3a8c1993aac4bd9aaf8", "messageHeadline": "Added explicit instructions to build for Jetson (#4714)", "author": {"name": "Nat Kershaw (MSFT)", "email": "nakersha@microsoft.com", "user": {"login": "natke"}}, "additions": 39, "deletions": 10}}, {"node": {"oid": "abbb7f6f5cfa69737c57363788f5ae712d3e1788", "messageHeadline": "Avoid duplicated calls of postprocess in training frontend (#4579)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 17, "deletions": 13}}, {"node": {"oid": "77c69a032538d1884ebfaa8892443589c64001fc", "messageHeadline": "Upgrade TensorRT to v7.1.3.4 (#4704)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 96, "deletions": 120}}, {"node": {"oid": "9c3153acd6d349cde598da49b43df1b9259f5ccf", "messageHeadline": "Improve shape inference for OneHot (#4452)", "author": {"name": "Oliver Rausch", "email": "Oliverrausch99@gmail.com", "user": {"login": "orausch"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "9c729d171952381e7750e19436a22976bb91dfc5", "messageHeadline": "Update notebook for mac since onnxruntime 1.3 or 1.4 in mac does not \u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 227, "deletions": 230}}, {"node": {"oid": "37c45c3d6bf4f1c6f152f22575b33edb3aac1aa6", "messageHeadline": "C# ResNet50 v2 sample/tutorial (#4722)", "author": {"name": "Marcus Turewicz", "email": "24448509+marcusturewicz@users.noreply.github.com", "user": {"login": "marcusturewicz"}}, "additions": 1278, "deletions": 0}}, {"node": {"oid": "61726e58f03c405c28680f14c850d9dc5630b930", "messageHeadline": "fix (#4697)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 5, "deletions": 6}}, {"node": {"oid": "c334b5738e9c7ab6ec7894034871eb29105ead4e", "messageHeadline": "Remove docstring for removed parameter (#4734)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 0, "deletions": 3}}, {"node": {"oid": "b22091dc911e04d3910fccbc667a7bc50bb8e2b3", "messageHeadline": "Add the framework to support prepack (#4413)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 535, "deletions": 136}}, {"node": {"oid": "33fe7700370f227b664319aba3616f39ad932be1", "messageHeadline": "Support log sigmoid gradient (#4719)", "author": {"name": "zhijxu-MS", "email": "43435212+zhijxu-MS@users.noreply.github.com", "user": {"login": "zhijxu-MS"}}, "additions": 45, "deletions": 0}}, {"node": {"oid": "7905c57f43eaccab9f7bf738d34c6c2dc8409082", "messageHeadline": "Revert \"Remove code which is not thread-safe. (#4454)\" (#4712)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 85, "deletions": 7}}, {"node": {"oid": "fc2f36c6083238e2a4b43f0f25414b5abb079f0c", "messageHeadline": "Shape independent gradient builder for Concat (#4675)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 331, "deletions": 65}}, {"node": {"oid": "8507bc1f480fc62e53970ec851c59243704ed47a", "messageHeadline": "[Android NNAPI EP] Enable test for BatchNormalization, enable dev_mod\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 126, "deletions": 58}}, {"node": {"oid": "4d39c6a6cb252b3baf382ae950e055c08bbc6db6", "messageHeadline": "Wire log(softmax) grad cuda kernel and add log(softmax) grad cpu kern\u2026", "author": {"name": "suffiank", "email": "sukha@microsoft.com", "user": null}, "additions": 229, "deletions": 33}}, {"node": {"oid": "9a73c8f448612ca7c5f0635f3f128c3809f63b86", "messageHeadline": "ReshapeGrad optimization (#4708)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 64, "deletions": 4}}, {"node": {"oid": "005fa5c3aec564ee14f78c261c0cb041a348792c", "messageHeadline": "Add initial Dockerfile for distributed training targets (#4578)", "author": {"name": "suffiank", "email": "sukha@microsoft.com", "user": null}, "additions": 403, "deletions": 1}}, {"node": {"oid": "e802b0498f3ccffda31a33a5ef591514c25b3cc9", "messageHeadline": "EnrichPyOpUT (#4681)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 115, "deletions": 5}}, {"node": {"oid": "43142a8225e62226d97e9f7174ef5090822c0448", "messageHeadline": "[Nuphar] added Gemm-to-MatMul conversion in model editor (#4691)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 606, "deletions": 23}}, {"node": {"oid": "5c5efa900d0b95fba8dbaf2a34e9b2267ade5d57", "messageHeadline": "Add .NET Core 3.0 nuget e2e pipeline tests (#4695)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 367, "deletions": 136}}, {"node": {"oid": "eb0f57f0e47120ce1d98797e2f0a1c15eb2cd050", "messageHeadline": "Localized Recompute for Gelu and AttentionDropout (#4402)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 256, "deletions": 75}}, {"node": {"oid": "0933148fc3a20ba2d9d12829790f2f3e92e5bcdb", "messageHeadline": "[Android NNAPI EP] change most of the exceptions to return Status (#4\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 852, "deletions": 662}}, {"node": {"oid": "1e054739b8eac6084ed033ac46ee8fcdcb1b75bb", "messageHeadline": "Remove the requirement of CUDA's version.txt (#4706)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "9d7284fc3bbaf473d103d759c4fa1a1e88693189", "messageHeadline": "Enable MatMul + Scale fusion (#4669)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 792, "deletions": 135}}, {"node": {"oid": "f9bd52f8525bf5279f2fb6990ae95665bb5ae2c7", "messageHeadline": "Log telemetry for WinML Native API for setting intra op num usage (#4\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 22, "deletions": 1}}, {"node": {"oid": "4bd9e8d05cf120f9e54dec251249a238e9be8798", "messageHeadline": "Stress-test and fix thread pool when work queues are full (#4690)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 62, "deletions": 4}}, {"node": {"oid": "d0297f8d249bc460c6c51b59eed8690853e9bca3", "messageHeadline": "Add 'Install ONNX' step to Windows GPU pipeline (#4696)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 20, "deletions": 2}}, {"node": {"oid": "49febba3c262054c2efe9df21d9a6f89ed7f6903", "messageHeadline": "Support int32 and int64 types for Tile CUDA kernel (#4684)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 71, "deletions": 55}}, {"node": {"oid": "e6ef3653a73a77e2dc015a308103be4fac435794", "messageHeadline": "Add Named Dimension Override API to LearningModelSessionOptions (WinM\u2026", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 93, "deletions": 2}}, {"node": {"oid": "bb9b452a88a2e48501008c023fc7de5847e99768", "messageHeadline": "resolves #3101 - fix nuget package restore for sdk-style projects (#4\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "0828a900e116ba04e5a2c6f838d6dcd8997cd1da", "messageHeadline": "Support easy way to request verbose logging in test runner (#4676)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "f74e55cfc984aead35c995b70afd7122d877bb21", "messageHeadline": "Merged PR 4986854: Opset 12: Clip, Max, Min, MaxPool, ReduceMax, Redu\u2026", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 30, "deletions": 4}}, {"node": {"oid": "01ca6392cb165f3ebad6fc3c28016adad5a26688", "messageHeadline": "Avoid building ONNX of every history ONNX versions in our CI (#4678)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 59, "deletions": 354}}, {"node": {"oid": "e70e9e2f6751e3ce14e7aeca1e21af46e5ad6350", "messageHeadline": "refine machine_info and output onnxruntime_tools version (#4679)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 34, "deletions": 8}}, {"node": {"oid": "6958f49dae3a26b3fc177fecc7aa87e177733d7e", "messageHeadline": "Added Dockerfile and build instructions for Jetson. Also set CUDA arc\u2026", "author": {"name": "Boris Fomitchev", "email": "borisfom@users.noreply.github.com", "user": {"login": "borisfom"}}, "additions": 98, "deletions": 32}}, {"node": {"oid": "b1bfff34e02c9176090469a23815e28e4f8a0a5c", "messageHeadline": "Support distill-bert fusion in transformers tool (#4631)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 143, "deletions": 74}}, {"node": {"oid": "8cf2c1c41075e455895f49e198bb3fe8e9e8231f", "messageHeadline": "Modify EmbedLayerNorm to support distill-bert  (#4666)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 138, "deletions": 45}}, {"node": {"oid": "1eadec0eea10ad0d0f92ea9df49d67207c8d0564", "messageHeadline": "Update Versioning.md for Windows 10 and Microsoft.AI.MachineLearning \u2026", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "c67dd693c271a1e31b354b0054532ea217db3aba", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2056, "deletions": 900}}, {"node": {"oid": "e9d20e9dba844aeed1d71c54e9219587f7b91c8c", "messageHeadline": "Revise Send and Recv (#4547)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 547, "deletions": 213}}, {"node": {"oid": "3588c5b5451adb2b1143b27fbf8c8150b35a6990", "messageHeadline": "Add GPT-2 test generation to convert_to_onnx.py (#4670)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 458, "deletions": 9}}, {"node": {"oid": "1fcd3eb376e68db16a9236afa9df9336f1e1edd0", "messageHeadline": "cancel night build on pyop (#4673)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "f9f25c5559a91507d0f0c4993f749c111c6b030e", "messageHeadline": "Remove featurizer from CI build (#4661)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 136}}, {"node": {"oid": "5ce675c3b99f38fcbbeaa60e7bc1636397a4e4be", "messageHeadline": "Expose Onnxruntime Intra Op thread controls through WinML Native API \u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 179, "deletions": 18}}, {"node": {"oid": "de0b04b971be68677dcfa686eeb32289530c1c72", "messageHeadline": "[Android NNAPI EP] Add support for dynamic output (#4650)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 471, "deletions": 328}}, {"node": {"oid": "282975aefbb2a644280785b09cc7d887a59ef17c", "messageHeadline": "[Android NNAPI EP] Add QLinearAdd op Support, move some throw with re\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 391, "deletions": 193}}, {"node": {"oid": "daa60c7dfbcf9021c9ba458460fd726258795504", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 295, "deletions": 226}}, {"node": {"oid": "51332e3c810dbcb22acfd941d9080578e6295d12", "messageHeadline": "Change Linux CI build time out value to 3 hours (#4664)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "319d30e50e0654b8a1769d2011adc84448457b13", "messageHeadline": "dnnl only run opset 10 model tests to reduce footprint/runtime (#4663)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0a8bfb10fa42a5f5b871bf28b0ff6694d2da866e", "messageHeadline": "Inbox WinML tests fail because Inbox loads binaries from system32 (#4\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 35, "deletions": 2}}, {"node": {"oid": "382f94c95cd7c991cc4385d9de313633a027c5ff", "messageHeadline": "Fix regression introduced in cast transformer (#4658)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 27, "deletions": 3}}, {"node": {"oid": "f90a2d46ae7a419dead0a6c0e16f6ccd27c03ecf", "messageHeadline": "Changes to support TNLRV3 fine-tuning (#4639)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 128, "deletions": 136}}, {"node": {"oid": "d8f3e46d458397b3fb335477d3efc91e22e160d8", "messageHeadline": "Readme updates (#4448)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 10, "deletions": 5}}, {"node": {"oid": "db475c4f358b0e936c8eb4b69d2eefa0b6be160d", "messageHeadline": "Add option for onnx_test_runner can pause after launch, make create_t\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 13, "deletions": 2}}, {"node": {"oid": "326cc686dfb138dc0c33c4f98ccecd0af4166d56", "messageHeadline": "Update notebook: disable GPU for tensorflow (#4649)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 80, "deletions": 76}}, {"node": {"oid": "6adc83d660e35ddfba8c19a78a87ce92d631e491", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1021, "deletions": 735}}, {"node": {"oid": "623dd53eb7daf79a4feefbb7980e66eaed964770", "messageHeadline": "Rename inner-scoped variable to avoid MSVC warning (#4587)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f3fc8ca954bc63d5c4739f18a2a2acb833526514", "messageHeadline": "Add input tensor calibration (#4619)", "author": {"name": "RRRachelllll555", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 40, "deletions": 26}}, {"node": {"oid": "d4983f83ff914a8b88170d4b2faf47e7c660bcf8", "messageHeadline": "Shape independent gradient builder for ops requiring broadcast  (#4586)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 722, "deletions": 428}}, {"node": {"oid": "948a33bdfcd0d37c217db2cc1059471ea6e22b7a", "messageHeadline": "FixPyOpSegFault&MakeItStaticLib (#4600)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 245, "deletions": 279}}, {"node": {"oid": "6c2bd127badbd8cdbc90663450b9110a080acb7e", "messageHeadline": "more types for comparison ops (#4634)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "72908b146eea343d2e7f6574bbe1da6a06e9b887", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 84, "deletions": 25}}, {"node": {"oid": "73ad92e7732944edf44b24a49efdf6e4c130a435", "messageHeadline": "Change ignore_index to 0 in Bert-Loss. (#4640)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a06cf6a3b361ec0003fe9151530bbddabccf6abe", "messageHeadline": "Show quantization model size in benchmark of transformer (#4626)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 19, "deletions": 5}}, {"node": {"oid": "73c99f8269dae08de9ee3af47a84514e8ae0c03d", "messageHeadline": "Set WINVER (#4636)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d73e01e5b90be91150a3551eb6fa963c30f9fcaa", "messageHeadline": "remove ENABLE_TELEMETRY macro (#4633)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "c08e5f55e9eabd4992f7ee5b33f70eb3560ad37f", "messageHeadline": "Fix misleading indentation (#4629)", "author": {"name": "albertomagni-ms", "email": "49027342+albertomagni-ms@users.noreply.github.com", "user": {"login": "alberto-magni"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "efa393e596e4141ef8f92a32e2536e689d12d45a", "messageHeadline": "WinML should dynamically link against onnxruntime.dll and only system\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 56, "deletions": 12}}, {"node": {"oid": "bee13d657b4ff36925b75d03652d69dce8088d02", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1151, "deletions": 126}}, {"node": {"oid": "222fd08f20765103e3bfbaf5de07b7f25373d516", "messageHeadline": "DirectML.dll is loaded via LoadLibraryW but should use LoadLibraryExA\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "1e67fff93cae31771c53e3c493358b28149edf24", "messageHeadline": "Add GetStringTensorElement, GetStringTensorElementLength and FillStri\u2026", "author": {"name": "Alisha Sonawalla", "email": "alsonawa@microsoft.com", "user": {"login": "alishasonawalla"}}, "additions": 143, "deletions": 8}}, {"node": {"oid": "c361a59cffec07eeef3464d9b3d8b1038482457e", "messageHeadline": "disable gpu timeouts in winml (#4604)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "48d969f4bf7e85380c08cdd42032446f105c66a9", "messageHeadline": "Constexpr CreateFeatureValueFromInspectable (#4460)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 18, "deletions": 17}}, {"node": {"oid": "9510f267448991b3fc1c6383a65c9847c46dcf3d", "messageHeadline": "[Python] Support more APIs for the SessionOptions class (#4596)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 111, "deletions": 85}}, {"node": {"oid": "9888c9e944da5bcd02c4d8578dc21b40e89afdc5", "messageHeadline": "SplitTraining op to support split as input  (#4597)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 521, "deletions": 2}}, {"node": {"oid": "aa328c2c20f685be5f5573198f56f5c9bbc7c862", "messageHeadline": "Update GratherGard to accumulate in fp32 (#4601)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "9c75c29403506a2473687c3ab24e3dba56ec7ee8", "messageHeadline": "refine opset version getter (#4602)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "5189530b7b5aef090dc647cfbc78074b7e0e7920", "messageHeadline": "Create the ConcatTraining op (#4595)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 336, "deletions": 7}}, {"node": {"oid": "15e1a1e1ace4faa49ec4e94f19af0e69a9c76acd", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 408, "deletions": 204}}, {"node": {"oid": "d5b98a13c2a7fd5635b85b5efb917f7af708d9df", "messageHeadline": "Move ReduceSumTraining op under orttraining (#4588)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 394, "deletions": 204}}, {"node": {"oid": "f35ce4677f74bf2840f15f78ecd638cdc2462798", "messageHeadline": "add double and uint8_t datatypes (#4603)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 14, "deletions": 0}}, {"node": {"oid": "de870c9637b18b9f0c97717945e311957625dc7a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 951, "deletions": 1251}}, {"node": {"oid": "ace41b80647a627805c766ef6b05e78a1e5cd206", "messageHeadline": "Force return_tuple=True to handle transformers breaking change of out\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "ea87c0d028ff14867994cb040b8b921993faaee1", "messageHeadline": "Update Transformer Optimizer documents (#4591)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 197, "deletions": 43}}, {"node": {"oid": "03ebe338508b2527713eac9f8fbb9fcb4f44e1da", "messageHeadline": "[Android NNAPI EP] Add support for LRN/Grouped Conv ops, fix issues w\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 243, "deletions": 99}}, {"node": {"oid": "c5df91874437bbbb4858b8e15806059c9dce773e", "messageHeadline": "improve calibration tool (#4561)", "author": {"name": "RRRachelllll555", "email": "35738743+YUNQIUGUO@users.noreply.github.com", "user": {"login": "YUNQIUGUO"}}, "additions": 315, "deletions": 391}}, {"node": {"oid": "1b253d18ef13b465a49230835405a92094a0443f", "messageHeadline": "Updated tags for MCR images (#4574)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "e2acb165e9acd20b4a6b09d72203ca35496aeed8", "messageHeadline": "Add exception check in training_runner when worker runs into error, a\u2026", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 112, "deletions": 74}}, {"node": {"oid": "c2ec3b734bb1816421fbf5dc6d8e162671aa0cd8", "messageHeadline": "[Android NNAPI EP] Remove dependency on external JD/DNNLibrary (#4576)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 40, "deletions": 613}}, {"node": {"oid": "3b53be001b274b6f40ffde4650f3fd2ad6a64d73", "messageHeadline": "Merged PR 4954300: Enable preliminary 8D support in DirectML EP", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "f0edd074fb6a957b7c40a5135af61654043ed1c2", "messageHeadline": "Optimize CreateEnv by not creating the logging manager instance if en\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 18, "deletions": 18}}, {"node": {"oid": "fa6d0350904da7d1496a7cdf85aef9ca9f7bf3f9", "messageHeadline": "Create WindowsAI zip files automatically as part of the pipeline (#4584)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "0b5117958aeb272678c53bd8d0824993e512e9c0", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 602, "deletions": 180}}, {"node": {"oid": "6b53a748678ab2995e22572b1a5d8c416b9e4fdf", "messageHeadline": "replace invalid sample (#4567)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "e00e2e84d6bf595715e766e5e4748e6ee1ac668e", "messageHeadline": "[OpenVINO] Update MCR dockerfile with Ubuntu 18.04 package names (#4581)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "f20afc4991568368a4919d0e383194a6481ffe1a", "messageHeadline": "Update ACL/ArmNN EP (#4571)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 584, "deletions": 173}}, {"node": {"oid": "822b23ff2fafdcbd2d1fdd0f4d7f8f6800ac622a", "messageHeadline": "Add support of EmbeddingLayerNorm (#4562)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "42d0ad8fec60a4db6d523fd72ec4b1dad28257bf", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 148, "deletions": 64}}, {"node": {"oid": "bf78e4d18b2f9a1ed6a91baa93f9d9d43578c832", "messageHeadline": "Handle cases which produce an empty output in the MatMul op family (#\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 148, "deletions": 64}}, {"node": {"oid": "8829d60f941d9f42f7201f4de2fe86fa12c4f0fe", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 838, "deletions": 74}}, {"node": {"oid": "0b4659c3fea7ef0131984bf98613e33954f9a774", "messageHeadline": "Populates TensorProto with tensor from protobuf file (#4535)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 11, "deletions": 10}}, {"node": {"oid": "603f2d113886632ae8f69bee6e1d7694869c451b", "messageHeadline": "Exclude a few OpenVino flaky tests (#4572)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "affdeb53c2b756ec49b0385d8d7e3e9d379993ec", "messageHeadline": "Add Python API for specifying device options. (#4205)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 526, "deletions": 7}}, {"node": {"oid": "e11629d9e43c299edbbf479e5a5d49100b06ff4e", "messageHeadline": "Revert \"Deprecate TrainableDropout (#4501)\" (#4564)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 292, "deletions": 48}}, {"node": {"oid": "d1500e1fc83933fac13d461e0c7c374f41263a84", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2694, "deletions": 357}}, {"node": {"oid": "ab4be8355f412c78197a575ca3568c6750f97d4d", "messageHeadline": "Add the BroadcastGradientArgs op (#4511)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 219, "deletions": 2}}, {"node": {"oid": "0008e92b4e1d7f1068d8638dfe5decaa5bfa5c4e", "messageHeadline": "Internal ReduceSum op that accepts axes as input  (#4522)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 370, "deletions": 10}}, {"node": {"oid": "e92e0860c885dfc66ed61fb0aee61c2aafa0f947", "messageHeadline": "BERT quantization notebook (#4543)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 962, "deletions": 0}}, {"node": {"oid": "ca0dd8246c132e315c0cbee8f0c428bc12329c8c", "messageHeadline": "NNAPI EP, add quantization support (#4530)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1031, "deletions": 230}}, {"node": {"oid": "7f9d9557b1f4a427dbffa7265af2f478034bcc2f", "messageHeadline": "Remove template<T> from RoiAlignBase (#4558)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 11, "deletions": 8}}, {"node": {"oid": "bbdabc2c48bf9ca3e5f34090f6ca97ed83be7fee", "messageHeadline": "Bump lodash from 4.17.15 to 4.17.19 in /nodejs", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "c2c4e6760b91f9ddca6aa6e6261ce48cd3deafe1", "messageHeadline": "Fix code sign validation errors in nuget and nodejs pipeline (#4527)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 54, "deletions": 14}}, {"node": {"oid": "1c5733ea9dcb1a1b261cf9bbb1d5ab3c514744af", "messageHeadline": "Silence binscope warnings (#4542)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9318b3a47f31d2d138374288a2f7483b6688e3c7", "messageHeadline": "Fix a typo in C++ sample application (#4556)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9d8023560718ebbdb42a52a3cda2b0782ab88cc7", "messageHeadline": "Reshape inputs for SoftmaxCrossEntropyLoss instead of transposing the\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 41, "deletions": 87}}, {"node": {"oid": "194367a2a370f3662f51eaa5053062dc42d95895", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 74, "deletions": 1244}}, {"node": {"oid": "bc1d197ddf81b06abee3707b1c7a942a4d2114f4", "messageHeadline": "Re-enable dnnl in CI build (#4544)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 168}}, {"node": {"oid": "0e91e45049744f0b3ce1b1701fd30503b3342e06", "messageHeadline": "[node.js binding] use official ORT C++ API (#4552)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1008}}, {"node": {"oid": "08235e16624299520c02e0032be9ad48002c2f72", "messageHeadline": "add Output() overloads (#4546)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 66, "deletions": 68}}, {"node": {"oid": "a15506ee848ed0b3c07086a5ac5b3edfd767760c", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 0, "deletions": 2}}, {"node": {"oid": "6c950a1df0833522b0240703981915ed485f0a55", "messageHeadline": "Exclude weight related types/shapes from bert loss. (#4548)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "e5716ebe4175d2c39cbb0f6ff7f305888171dc7e", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2193, "deletions": 1646}}, {"node": {"oid": "b42fe49f564070275f6b138695433ce2b164f0be", "messageHeadline": "Add a script to convert gpt2 to onnx (#4541)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1201, "deletions": 947}}, {"node": {"oid": "0229a6a929587e7e48238571eba31370f162a305", "messageHeadline": "[C++ API] add SessionOptions::SetLogSeverityLevel() (#4545)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "6eb5549cb976e34efd75e32c2e20e21beddf01c8", "messageHeadline": "Deprecate TrainableDropout (#4501)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 48, "deletions": 292}}, {"node": {"oid": "fdc5c308c4a579bcf96a170cffa3c00e761eb8cd", "messageHeadline": "introduce macro ORT_API_MANUAL_INIT in C++ API (#4536)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 92, "deletions": 91}}, {"node": {"oid": "21d2728974bad57dcbed6b5bffcc250d6c2ee0da", "messageHeadline": "Revise pipeline schedule to consider communication ops (#4524)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 844, "deletions": 315}}, {"node": {"oid": "5b7050321feb4f09f0dfadc380cf9bcf7f827e39", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 306, "deletions": 165}}, {"node": {"oid": "1154b6e57919d50b95d98ca34aa63496d26b10e5", "messageHeadline": "Merged PR 4933520: (ap build) Update to GSL 3.1 - Fix signed unsigned\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "183098e344a70efa612e8e5efb45b0d83eb086b3", "messageHeadline": "adding concat logic when initial path is empty (#4525)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 70, "deletions": 50}}, {"node": {"oid": "d1f45f93611714ca652472d2333b40511cc682c1", "messageHeadline": "Handle nodes with dtype attribute causing diffs in inferred and actua\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 59, "deletions": 6}}, {"node": {"oid": "8b86c5cdb5c2f0ee3bcd5217e19bb16f63d93636", "messageHeadline": "Merge int32/uint32 and int64/uint64 MatMul kernels (#4531)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 27, "deletions": 52}}, {"node": {"oid": "02aea5d2d4f0a3a042158ccca97bb5fe41378b4b", "messageHeadline": "rename telemetry provider back to Microsoft.Windows.AI.MachineLearnin\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5086e55a35f83e3137bdb34b6d7210c97a512e6a", "messageHeadline": "Fix condition of running tests in win CI (#4459)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 10, "deletions": 9}}, {"node": {"oid": "2189c77e5b77f07aa2657b3b6978eeb87d4d5f95", "messageHeadline": "static_typename (#4520)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "b43ce2d7adf21c63cd784cd7d3da63f24273787e", "messageHeadline": "Replace loss function in BERT_LOSS with SoftmaxCrossEntropyLoss. (#4509)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 105, "deletions": 27}}, {"node": {"oid": "76b31d6ce2a1fade1746f340cee32ca0e9bb4dfa", "messageHeadline": "fix xcode alerts (#4470)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 22, "deletions": 20}}, {"node": {"oid": "42a5a4bff124c33a87cfafcab9c1653e3d9da4eb", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1039, "deletions": 203}}, {"node": {"oid": "8ada440961a9771185d4c57c71f87dac3fd62a63", "messageHeadline": "Move model tests to onnxruntime_test_all (#4521)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1038, "deletions": 202}}, {"node": {"oid": "5f188f4cf49d8a5cfb78b8785449e14ac81f00f4", "messageHeadline": "ci fix (#4519)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "938a4a7fa2db965e993202a15fd5865bb22a8dda", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1489, "deletions": 565}}, {"node": {"oid": "0ebe2fab51e8ec35c4682deb2adb6ba88f2fa618", "messageHeadline": "Refactor TensorRT EP code to better handle dynamic shape subgraphs (#\u2026", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 546, "deletions": 247}}, {"node": {"oid": "cf92497c160b3a72ae663ab4dd9cd818331b62db", "messageHeadline": "Nnapi, add auto_pad support for Conv/GlobalAveragePool/AveragePool/Gl\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 416, "deletions": 314}}, {"node": {"oid": "34f73fa1aa8740e0869c89d24d3fd1220222198b", "messageHeadline": "Add sudo --preserve-env option to allow environment to go through to \u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f721f5f1cd2a292e1197393a080c60b13549aa61", "messageHeadline": "Liqun/multiple choice (#4480)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 526, "deletions": 3}}, {"node": {"oid": "00d775e5674c0b14d531c6bd4aa9184e8425bbed", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 863, "deletions": 357}}, {"node": {"oid": "7d96960ec83abc65c7c4b8fc085ea65a7a1fa237", "messageHeadline": "support pipeline partition with shared initializer (#4321)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 739, "deletions": 306}}, {"node": {"oid": "1ebe598286bd12d544325b4c50ae9f68490c650f", "messageHeadline": "Conditionally compile without std::is_trivially_copyable to satisfy o\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "ee5ca27ae2b9d3994bc95287facf51d45b9fd180", "messageHeadline": "Split Microsoft.AI.MachineLearning.nupkg in a NuGet package and symbo\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 117, "deletions": 50}}, {"node": {"oid": "0714c1db984aa7479f53019cc15bd702b2581818", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 109, "deletions": 113}}, {"node": {"oid": "273a4ff1b3f3b948fd9a677e018e13307aa420ed", "messageHeadline": "Merge branch 'DmlDev' of https://microsoft.visualstudio.com/DefaultCo\u2026", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "25885cf7d0240e60e0bd4beae82825048e95585b", "messageHeadline": "Add option --torchscript in benchmark_gpt2.py (#4500)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 109, "deletions": 113}}, {"node": {"oid": "907952ef8ce47f4fa420b4ffef3651fa24c4f99e", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 177, "deletions": 101}}, {"node": {"oid": "a2838d71a0b0a220fab4f58cc8cab8d81756f31e", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 177, "deletions": 101}}, {"node": {"oid": "9d10a2d589c0e0f723bdc72846fd5d1da45ce8b7", "messageHeadline": "Merge branch 'DmlDev' of https://microsoft.visualstudio.com/DefaultCo\u2026", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 789, "deletions": 710}}, {"node": {"oid": "a95ae164f7e21c2f31f4748e127b52c075c6ecb3", "messageHeadline": "Create N-1 threads in intra-op pool, given main thread now active (#4\u2026", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 109, "deletions": 54}}, {"node": {"oid": "0bff55512e3a8ae016d06d667e3b378566534327", "messageHeadline": "updated expected values for frontend test to pass frontend e2e pipeli\u2026", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 68, "deletions": 47}}, {"node": {"oid": "7753b8dce68bed48d3835e78736f8aeca8ccd582", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 8749, "deletions": 2704}}, {"node": {"oid": "e0eddf502c31e27b45fcec2544bf438f0229430a", "messageHeadline": "Bump version to 1.4.0 (#4496)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "3d4ac851247136183bd01e6e9aa66e5a7e486ad1", "messageHeadline": "Add quantization benchmark for transformer based model (#4482)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 203, "deletions": 79}}, {"node": {"oid": "a3c358fd29cb856049496f334708b70c633acc36", "messageHeadline": "Split the shared ComputePadAndOutputShape into 2 separated functions \u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 136, "deletions": 174}}, {"node": {"oid": "3441c687b7e3f694d842cd83d6523d447c4183ca", "messageHeadline": "Revert \"Remove docstrigs if __ONNX_NO_DOC_STRINGS\" (#4495)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 429, "deletions": 432}}, {"node": {"oid": "5f8f443ac49cbe2b14702629da66b7fe5d7b5678", "messageHeadline": "Android CI build, test copy, emulator boot improvement (#4481)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 10, "deletions": 20}}, {"node": {"oid": "b1b75b35ac636a37afaf66bdff2f070e8cc3fe00", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 8866, "deletions": 2900}}, {"node": {"oid": "35ee00d888790b1d8ac390ccc06b8b177e3e4c80", "messageHeadline": "Pin typing version. (#4490)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "07455cff289d411a3dd0bc58c728d4cc015249ed", "messageHeadline": "Support double type for Greater CPU (#4373)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "f18dee84c24527c78f1d8e38f4408d60f74cae37", "messageHeadline": "Remove docstrigs if __ONNX_NO_DOC_STRINGS (#4494)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 432, "deletions": 429}}, {"node": {"oid": "c71c49aaa03a07ee4c05ffb6a9af3f150bce3c1d", "messageHeadline": "Make TArray safer to use and update method name for consistency. (#4483)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 174, "deletions": 172}}, {"node": {"oid": "00706e1502049610c93425308d9603f64be0b952", "messageHeadline": "dont add deps for uwp apps (#4485)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "3e48ffd21c4c924010b1a50a6de785a9486a790b", "messageHeadline": "Move AutoPadType to common.h (#4474)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 81, "deletions": 89}}, {"node": {"oid": "e96a829e8458528c9569aa6fd44939a11d642fc8", "messageHeadline": "Handle multiple embed nodes in transformer optimizer (#4471)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 234, "deletions": 111}}, {"node": {"oid": "6a9a9a35be7a87109a7ab77cfa2db3695a2b2581", "messageHeadline": "fix crashes caused by test runner (#4475)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 61, "deletions": 33}}, {"node": {"oid": "26ebcfab88bcdea1aa567729e89e865c05aa2526", "messageHeadline": "Fix Nuget GPU pipeline (#4462)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 21, "deletions": 3}}, {"node": {"oid": "9b4c54bcef2a6f2d590f0db6ceacfbac7d524241", "messageHeadline": "Enable onnxruntime_test_all for NNAPI EP (#4476)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 76, "deletions": 54}}, {"node": {"oid": "6c7da5e9d3893b3f2db3ba684ab775884122f586", "messageHeadline": "Optimize CUDA Sum op kernel and refactor CUDA elementwise variadic in\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 739, "deletions": 297}}, {"node": {"oid": "04586fc09d5f8cc37b75eb65d7e4d33ad4461005", "messageHeadline": "Fix segmentation fault caused by invalid tensor type (#4467)", "author": {"name": "Prabhat", "email": "prabhatr@usc.edu", "user": {"login": "prabhat00155"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "ccbf49e59f6bef897a94595e2213263b37d64ff3", "messageHeadline": "Fix avx2 load 32 bytes buffer overrun. (#4455)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 344, "deletions": 169}}, {"node": {"oid": "d4db83858b9c8f7c8cd5952a375e19fe16ba2544", "messageHeadline": "Only quantize gather with initializer (#4469)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "bec18eb3f41ead8a27eb3ddfb9a7fe6cd528e29a", "messageHeadline": "[Node.js binding] support CentOS 7 in CI (#4447)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "ca5af9d622b3f9a63092956c27279587c67e9967", "messageHeadline": "Add modern C++ standards for Ort::Value (#4367)", "author": {"name": "Josh Bradley", "email": "jgbradley1@gmail.com", "user": {"login": "jgbradley1"}}, "additions": 46, "deletions": 11}}, {"node": {"oid": "7fb194d03db43624bee2afb6a6a23bf98135d00f", "messageHeadline": "Update convergence baseline for ci_test. (#4465)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "3effac2990fe73a6725d27748ba52f19f7b70e8c", "messageHeadline": "Experimental C++ API examples (#4358)", "author": {"name": "Josh Bradley", "email": "jgbradley1@gmail.com", "user": {"login": "jgbradley1"}}, "additions": 249, "deletions": 8}}, {"node": {"oid": "5dc7339be68661b335bbc0cef01a0a064e6dc2fc", "messageHeadline": "Add quantization tool to python package (#4458)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 51, "deletions": 0}}, {"node": {"oid": "0ca4f7eb30e22670e3606510c0bb00cc051455b6", "messageHeadline": "Update Git submodule cgmanifests. (#4461)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "f24d8e4587dc49b9c9c2632251c5c9e06d7ba01b", "messageHeadline": "fix build break from PR#2850 api change (#4451)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 16, "deletions": 24}}, {"node": {"oid": "cb5c4292b8938244cc59843f24eb2f57e73f4940", "messageHeadline": "GPT-2 Attention Fusion without input mask (#4456)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 95, "deletions": 17}}, {"node": {"oid": "5222b2c6c0366033aaa78b95b2a0a49984984203", "messageHeadline": "Remove code which is not thread-safe. (#4454)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 6, "deletions": 60}}, {"node": {"oid": "05757b4c3c44f7b1ee8b295f5e6ca2a66389c5bd", "messageHeadline": "Transformer benchmark: add option to use raw attention mask (#4446)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 105, "deletions": 43}}, {"node": {"oid": "b156ae4448fdeeccc0bf66b8491be75fe9b368aa", "messageHeadline": "Support training_mode flag in eval (#4324)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 123, "deletions": 52}}, {"node": {"oid": "71aec2adcb015b29c58f984f25a6c98d02d7b550", "messageHeadline": "Custom op export test template (#4383)", "author": {"name": "Negin Raoof", "email": "neginmr@utexas.edu", "user": {"login": "neginraoof"}}, "additions": 43, "deletions": 20}}, {"node": {"oid": "063156d98d7f770dd4f82ca37edf9c76f2054426", "messageHeadline": "IOBinding docs (#4432)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 53, "deletions": 0}}, {"node": {"oid": "6d6b6b54a544f69fb201a9aab7097358f6a440ff", "messageHeadline": "Support binding a graph output to a specific device via the Python bi\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 193, "deletions": 66}}, {"node": {"oid": "aa06d308a6e972bc36f2738b010825859f2ce862", "messageHeadline": "Build new AVX file with /ARCH:AVX (#4442)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "e62686c36e6295a558b25052a9cc775baa3b0b29", "messageHeadline": "Remove use of RTTI in CUDA provider (#4444)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "fdb4a3a2e887aeea235e6750add70202e307ebde", "messageHeadline": "Add cppwinrt and cswinrt tests in windowsai nuget pipeline (#4381)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 640, "deletions": 1}}, {"node": {"oid": "612f52c9750fd20e05d0ce9941281dbb1d7e4d16", "messageHeadline": "add bias for DynamicQuantizeMatmul (#4440)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1f1384f8a9d0b85ea03fae9b0febfdbabe0ca0e1", "messageHeadline": "Update dependency introduced by fuzzing change. (#4438)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 216, "deletions": 0}}, {"node": {"oid": "eabf6dc9eebffaf4aa6b2ae2191453b0540bb29e", "messageHeadline": "Add Fusion for GPT Attention with both past state and attention mask \u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1204, "deletions": 206}}, {"node": {"oid": "7baf3749390589d72d9ff962059a76fbb6785d00", "messageHeadline": " Change the input to NNAPI EP ModelBuilder from ModelProto to GraphVi\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 667, "deletions": 724}}, {"node": {"oid": "632b2896f374ce664ca1be363795427c9dc1f00f", "messageHeadline": "Onnxruntime fuzzing (#4341)", "author": {"name": "EronsJ", "email": "EronsJ@users.noreply.github.com", "user": {"login": "EronsJ"}}, "additions": 1699, "deletions": 24}}, {"node": {"oid": "ec35a1b51423d4622010c386c14a12a7c9cbc210", "messageHeadline": "Remove unused initializer in graph after embed fusion (#4436)", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "3ef449816c2c01703d6aa9d99f15bd77f1325515", "messageHeadline": "MLAS: support prepacking APIs for quantized GEMM (#4433)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1196, "deletions": 269}}, {"node": {"oid": "dd73e8c016a7b8037818d8f54aafb2df2ffb13b5", "messageHeadline": "add function initialization back to graph resolve (#4434)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 2, "deletions": 7}}, {"node": {"oid": "0fdb1e9f6005aa3441bd061aec3dc5b81c69dc22", "messageHeadline": "Liqun/roberta (#4408)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 78, "deletions": 21}}, {"node": {"oid": "358848433699f6bc834f53392f8b1f1ee2c24633", "messageHeadline": "use system libnsync (#4377)", "author": {"name": "Christian Goll", "email": "Christian.Goll@gmail.com", "user": {"login": "mslacken"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a44eb7dd08a2eddd0f3c165ca9db194563157316", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 3723, "deletions": 912}}, {"node": {"oid": "77cf51b13c265d3fffcb14d010512c3e50e98048", "messageHeadline": "Fix symbolic_shape_infer for Resize with roi (#4426)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 8, "deletions": 4}}, {"node": {"oid": "0d4a65eede737b10cbbdce563b413199b29c06c2", "messageHeadline": "Fix Vitis-AI EP for memory info into IAllocator move (#4404)", "author": {"name": "jornt-xilinx", "email": "65705538+jornt-xilinx@users.noreply.github.com", "user": {"login": "jornt-xilinx"}}, "additions": 4, "deletions": 7}}, {"node": {"oid": "8bcdefc9c1c5406e1f6459b6d32ff032e3c064e5", "messageHeadline": "Optimize GatherND (#4097)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1091, "deletions": 131}}, {"node": {"oid": "bd11ab68161bcb1cd3af7881fc1875eb3f9929c8", "messageHeadline": "Optimize LayernormGrad (#4156)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 593, "deletions": 83}}, {"node": {"oid": "33e06be4ac13c6a58d3365f4cc6e33ef5ed410b4", "messageHeadline": "optimize transpose CUDA kernel (#4233)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 296, "deletions": 23}}, {"node": {"oid": "dba22b17b47656a82752a40960964794de652dbf", "messageHeadline": "Update BiasGeluGradDxKernel and tests. (#4400)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 47, "deletions": 19}}, {"node": {"oid": "93d49647277c13e96084a54d6e40b367373d345f", "messageHeadline": "Use single OpKernel for u8u8 and u8s8 types (#4414)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 333, "deletions": 589}}, {"node": {"oid": "4df8a1e2402d22d441539e48c322468ca250b8d7", "messageHeadline": "Use the file size while reading onnx models. Ensure models are loaded\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 60, "deletions": 22}}, {"node": {"oid": "d22f6fddf72cd984553b0331a23fbf89b215c7cd", "messageHeadline": "Add ability to specify just the device when using IOBinding for an ou\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 77, "deletions": 31}}, {"node": {"oid": "28e4c0edf5b9a2ef18be6c1e19e6b2772ec18ff2", "messageHeadline": "Keep loss_scale and Whole Loss Subgraph in FP32 during Mixed Precisio\u2026", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 317, "deletions": 55}}, {"node": {"oid": "7a05b3ca87ee32b3fca01f9faac4e11991a1739d", "messageHeadline": "Increase python packaging pipeline timeout (#4412)", "author": {"name": "suffiank", "email": "sukha@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "67a7d93b4901dc45873090a52f490990efa2cbb4", "messageHeadline": "Fuse MatMulInteger and scale followed (#4350)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 997, "deletions": 54}}, {"node": {"oid": "10c25416bb68f32c475f1a9ed827e75f7dea3c21", "messageHeadline": "Remove use of RTTI in CUDA provider (#4410)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "eabc1616e60ca41ba94bd282f9f8f8e083069aba", "messageHeadline": "Rename variable in InferenceSession class so as to not clash with an \u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 14, "deletions": 13}}, {"node": {"oid": "f6bf66c8cf7fc57d849642d99d13058ba62c7f02", "messageHeadline": "Adjustments to MPI and NCCL library discovery on build (#4407)", "author": {"name": "suffiank", "email": "sukha@microsoft.com", "user": null}, "additions": 6, "deletions": 1}}, {"node": {"oid": "ff6437136793f7bd52de91e526a83470442a49e5", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f4e0070c2eb56209c9211d0d9a930752789426ff", "messageHeadline": "Bump mysql-connector-java from 8.0.15 to 8.0.16 in /tools/perf_util (\u2026", "author": {"name": "dependabot[bot]", "email": "49699333+dependabot[bot]@users.noreply.github.com", "user": {"login": "dependabot[bot]"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "84a34a38d0ff84890d32b84652471278cbb44c41", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 57, "deletions": 55}}, {"node": {"oid": "0bef9d51141f05cb432187fff71251b0f0887a90", "messageHeadline": "Fix the broken Android NNAPI CI (#4403)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 7, "deletions": 13}}, {"node": {"oid": "ef602835b018a3437a8d0f433e77d9264dd7c3d2", "messageHeadline": "update getfunctionbody (#4396)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3bb6a865ccb1ffd011c03d1ba28179988f0850f6", "messageHeadline": "Revert \"remove openmp and scipy from build pipelines (#4305)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 48, "deletions": 40}}, {"node": {"oid": "5cb5081c8fc0b6c4cbeb8400b7854e657a479ec4", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 431, "deletions": 1195}}, {"node": {"oid": "4c0236d6c1503d9df1df24cfa456276b853e9f8d", "messageHeadline": "Update MCR container instructions with dynamic device selection info \u2026", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "5c23b1719693d6d9ec2e8f532209aee265393bb4", "messageHeadline": "MLAS: more prepacking kernel changes (#4397)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 390, "deletions": 408}}, {"node": {"oid": "d4341ea2de8c2ada0ddbfbf4acd577851d9cf897", "messageHeadline": "Merged PR 4870266: Refactor fused graph kernel so dmlxp and ort share\u2026", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 412, "deletions": 279}}, {"node": {"oid": "2d54c89d77424fe1bb4bb0b5ff05c9e683175812", "messageHeadline": "Update filename and Cleanup unused cudnn kernels (#4387)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 21, "deletions": 782}}, {"node": {"oid": "010445fc527e012b4c8b6eb294064b251c85bae3", "messageHeadline": "handle Floor and SplitToSequence (#4384)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 18, "deletions": 5}}, {"node": {"oid": "6ee6fdbbba2b83dcaf3a911afa9be1c5098ccb0c", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 19, "deletions": 24}}, {"node": {"oid": "473cd5545fde1d9cbcf48b7af898c5837f6e3e80", "messageHeadline": "Simple support of MatMul U8S8 on ARM to pass tests (#4392)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 19, "deletions": 24}}, {"node": {"oid": "bffbae33986fd57a3bba0181db42a85bec9a54e9", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 1919, "deletions": 92}}, {"node": {"oid": "7ec9a732021646d469eb5cb458d15b2651043bbd", "messageHeadline": "deprecate frontend layernorm postpass (#4372)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 12, "deletions": 11}}, {"node": {"oid": "7fea332f932c71d2f2f57bdc527237fb12b1591b", "messageHeadline": "Support builds without RTTI (#4333)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 59, "deletions": 27}}, {"node": {"oid": "5dcb9b485825141704ceb9a9f234fb11d397b992", "messageHeadline": "Liqun/backprop deterministic graph (#4315)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 72, "deletions": 45}}, {"node": {"oid": "94c98aa0a77597babf981b519a96cc810122055c", "messageHeadline": "qlinaradd for arm/sse2/avx2 using intrinsic, enable binary broadcasti\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1774, "deletions": 3}}, {"node": {"oid": "49268c42da256cf9589dc8793f6918df4e84f95d", "messageHeadline": "Change the way java home is set on Mac OS for CI and Java publishing \u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 6}}, {"node": {"oid": "1b26dfc8ace8b515b5319e4bbb7abafa8ee39122", "messageHeadline": "Merged PR 4868876: Merge Onnxruntime github with dmldev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "531126916b0ea3bd2fd5f50d66f932649a12d384", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into HEAD", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "4051667d01591f765649444c99d73aed14a0c521", "messageHeadline": "Merged PR 4868833: FIx merge conflict", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "f422bcb3722b65cb738884a1ee9b495fabc840b4", "messageHeadline": "Merged PR 4868758: Fix merge conflict in merge of Onnxruntime to DmlDev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 1316, "deletions": 483}}, {"node": {"oid": "018269f29f4ea0134914a0da66aebc7850d5900e", "messageHeadline": "Merged PR 4868304: Manual merge of Onnxruntime github into DmlDev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 1829, "deletions": 529}}, {"node": {"oid": "6365760906ebf7f2198a48b6612440b333dbd32e", "messageHeadline": "BiasDropoutFusion  (#4167)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 1028, "deletions": 318}}, {"node": {"oid": "0404763f23e4d2e1b06128b3545ad7c1042a7a66", "messageHeadline": "Update function body initialization for ONNX functions (#4332)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 68, "deletions": 27}}, {"node": {"oid": "7cb2c3f025d1c2ae894d90c16f3dd0c75480cac9", "messageHeadline": "Merged PR 4852260: DML EP remove redundant rank checks for higher dim\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 11, "deletions": 16}}, {"node": {"oid": "37b624b68887158636a308eb366f2a89cd4b0601", "messageHeadline": "Match More EmbedLayerNormalization Patterns for Bert Model Graph Fusi\u2026", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 220, "deletions": 138}}, {"node": {"oid": "755675541a2572f7feef27ce515673a6c478134d", "messageHeadline": "NCHWc + Sigmoid optimization (#4360)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 40, "deletions": 2}}, {"node": {"oid": "4380b8ba681647786080ffc329c8dd6c189e20ac", "messageHeadline": "adjust bs size (#4375)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "89c6da99b5d229d67b465a2ecf158a4254932cd3", "messageHeadline": "fix output shape calc for matmul (#4362)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "a4127fc185c25dba0c918ba9e0996805ad35a92e", "messageHeadline": "Add stale bot (#4323)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 22, "deletions": 0}}, {"node": {"oid": "55f25a4bbf67fd26b091ef2758ddfbd0404f0d8e", "messageHeadline": "Update Attention op to support attention mask for GPT-2 (#4330)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 967, "deletions": 214}}, {"node": {"oid": "2601f8e1b48bc8783d150cc6b7389bc357dda101", "messageHeadline": "Support to build CUDA EP for NV Ampere GPU (#4345)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 7, "deletions": 5}}, {"node": {"oid": "465140b3842289ddc86d908bffe47c0701037059", "messageHeadline": "Misc fixes to Conv and ConvTranspose CUDA kernels (#4281)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 63, "deletions": 48}}, {"node": {"oid": "35a048ef9b9e840abde5e04cf335af09aaac8729", "messageHeadline": "Ignore one failed test in DML (#4366)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5f4e63ede66eced1144e1019cd02b6e3d09b3470", "messageHeadline": "Add nhwc support for NNAPI EP, add concat op, handle concurrent calls\u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 633, "deletions": 177}}, {"node": {"oid": "88402f52934ac4a8458e064cbb0f61507e9cf914", "messageHeadline": "Make DML operator registration constexpr (#4219)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 94, "deletions": 87}}, {"node": {"oid": "ee48b89350011a07d846b6d39a7c57069d135489", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 554, "deletions": 513}}, {"node": {"oid": "9d1c86f7a5dfa4f2d4dd4cc24ede32c9921e34ee", "messageHeadline": "Merge branch 'DmlDev' of https://microsoft.visualstudio.com/DefaultCo\u2026", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "012aaa649194ce78e3493837a7fb7ebcfee4bf76", "messageHeadline": "Minor optimization in CUDA Reduction ops (#4353)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 16, "deletions": 17}}, {"node": {"oid": "274e6b41535f79808e2c6d52406851479ab6dca0", "messageHeadline": "Cleanup SessionState. Move allocator lookup to SessionState. (#4194)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 538, "deletions": 496}}, {"node": {"oid": "386046426bdfd2f64be648405e909c921b8f103b", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 10258, "deletions": 1694}}, {"node": {"oid": "8b0968b622bd772147e9e6b172aab9319646a754", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 10258, "deletions": 1694}}, {"node": {"oid": "4a1ecd98798f24a2564b5eeb66b141aaac6d5cb7", "messageHeadline": "[OpenVINO] Upgrade OpenVINO docker base to Ubuntu 18.04 (#4346)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "d1777910a8afac133b339225eb167b3aef0636e7", "messageHeadline": "fix onnx server build failure. (#4347)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c3c4ce5cebba6bb4d6e1745cbf7d5a51cad1067e", "messageHeadline": "refactor prototypes into headers (#4337)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 10, "deletions": 8}}, {"node": {"oid": "fc5e65a22dcd263301522f289a3d1b852f001888", "messageHeadline": "Add quantization support for GPT2 past state and use model to generat\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 594, "deletions": 628}}, {"node": {"oid": "ceedf126a2819e9f0ca488680efba86242ebd536", "messageHeadline": "[nGraph] Deprecation notice for nGraph EP (#4344)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 34, "deletions": 0}}, {"node": {"oid": "381f4c442a6131bed5a8eff1951f3b2a865d93cd", "messageHeadline": "LayerNormFusion - Cast support (#4320)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 85, "deletions": 11}}, {"node": {"oid": "9e0f5fc7af62a6b58fa6296ae6648b02fdafd8f7", "messageHeadline": "The initial PR for NNAPI EP (#4287)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 5727, "deletions": 20}}, {"node": {"oid": "37cbe8551da26c7ceb3f4ec1d77149314493c0a0", "messageHeadline": "Adding export registration and tests for custom ops (#4248)", "author": {"name": "Negin Raoof", "email": "neginmr@utexas.edu", "user": {"login": "neginraoof"}}, "additions": 369, "deletions": 0}}, {"node": {"oid": "990b43ddf2527a95792b1621f4135359d16b6c96", "messageHeadline": "Add modern C++ standards to the C++ API (#4217)", "author": {"name": "Josh Bradley", "email": "jgbradley1@gmail.com", "user": {"login": "jgbradley1"}}, "additions": 144, "deletions": 1}}, {"node": {"oid": "72fb5183d4d98de899a353164f99d131fdf3742d", "messageHeadline": "Fix Windows ARM64 break (#4343)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 13, "deletions": 10}}, {"node": {"oid": "a37e2e33b4ce93a8a4124be932298dc338d97071", "messageHeadline": "Add compatibility with Protobuf 3.12 (#4291)", "author": {"name": "Chih-Hsuan Yen", "email": "yan12125@gmail.com", "user": null}, "additions": 70, "deletions": 62}}, {"node": {"oid": "5db67ec0005046047ab119405fc4820b74b9b060", "messageHeadline": "Fix python package issue and upgrade the linux image to 2010 (#4342)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "bfc888613f4e831d29c8b0bc17182ae061712553", "messageHeadline": "Migraphx improvements (#4328)", "author": {"name": "Shucai Xiao", "email": "shucai.xiao@amd.com", "user": {"login": "scxiao"}}, "additions": 270, "deletions": 144}}, {"node": {"oid": "0b450dcd9f34813912a1f66595598a68878b6b7e", "messageHeadline": "Enable BiasGelu fusion for training (#4146)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 815, "deletions": 203}}, {"node": {"oid": "c0a3320522c8d83ec28eb5daba3c8f0ea2602b38", "messageHeadline": "Merged PR 4851009: move clamp_cast under OperatorHelper namespace", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "b544f5c83c7c4094a78917efc3293566080036f6", "messageHeadline": "Sample updates (#4303)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 60, "deletions": 38}}, {"node": {"oid": "645a988c04b872b2efeb596d3e0a65c06e0985c1", "messageHeadline": "Support binding input only for IOBinding in python api. (#4079)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 188, "deletions": 17}}, {"node": {"oid": "a08805daf954348262fe6fd831ce23a0a3f593e7", "messageHeadline": "Fix a minor typon in POM file name (#4250)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3fc68cb150c38897f91d43933622d793d4a18076", "messageHeadline": "Remove non-trivially-destructible thread-local from thread pool state\u2026", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 52, "deletions": 39}}, {"node": {"oid": "a3b466cdf1fefdc9097d008f69fd6a9ceeed84ab", "messageHeadline": "fix python ep default ordering.  (#4335)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 11, "deletions": 7}}, {"node": {"oid": "151ef1c8a51875324d48501fd6362aaf86be9d8d", "messageHeadline": "Add C++ wrapper for GetAvailableProviders() C API (#4313)", "author": {"name": "Prabhat", "email": "prabhatr@usc.edu", "user": {"login": "prabhat00155"}}, "additions": 22, "deletions": 2}}, {"node": {"oid": "a6d10376dfc7b4613ecf59d0b02638d7bba42288", "messageHeadline": "Fix build error when USE_NCCL is defined. (#4334)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "0d9db2b28d7ca72865405d7fbe47c6221f01d4ab", "messageHeadline": "add informative error message regarding symbolic dimensions (#4297)", "author": {"name": "Josh Bradley", "email": "jgbradley1@gmail.com", "user": {"login": "jgbradley1"}}, "additions": 14, "deletions": 12}}, {"node": {"oid": "64264c3846065605199dd7eaa05a8ae1704567cb", "messageHeadline": "Allow --cmake_generator to work on macOS (#4278)", "author": {"name": "Aaron Bockover", "email": "abock@microsoft.com", "user": {"login": "abock"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "15c07c75f87327cfd8354bdc26f97367366c84e2", "messageHeadline": "[OpenVINO-EP] Upgrade version info to 2020.3 in docs (#4304)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 13, "deletions": 8}}, {"node": {"oid": "a241eb0bbe44636c8656af17001470c39346f9c5", "messageHeadline": "Renaming --partition_optimizer to --deepspeed_zero_stage (#4312)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 50, "deletions": 31}}, {"node": {"oid": "7e71ff2a1f1f46c89884f4f4f3c056d8ce275868", "messageHeadline": "Match Reshape Subgraph Pattern For GPT2 (#4279)", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 310, "deletions": 34}}, {"node": {"oid": "5c6a27408a8e2a2adade0f7dd849cf052d7f9012", "messageHeadline": "Remove signed/unsigned compiler warnings, add additional pipeline tes\u2026", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 35, "deletions": 12}}, {"node": {"oid": "44f06ec480878df105e39d422f889e483d79e746", "messageHeadline": "Fix memory usage when loading a model + some other minor fixes to avo\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 6}}, {"node": {"oid": "5dd3ebb3b146ce001327c154258760cf919e0d49", "messageHeadline": "Tune setting for when to use MlasComputeSoftmax due to changes in #39\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 8}}, {"node": {"oid": "f26c149d7dab96da24f516cb9f785f2d2c656e0a", "messageHeadline": "Set NonZero Output Shape for Gradient Building. (#4246)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 122, "deletions": 0}}, {"node": {"oid": "20e205aa0aa5a8eeaad5de569d9ca3845c857921", "messageHeadline": "[OpenVINO-EP] Changed the default scheduler for VAD-M (#4295)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 154, "deletions": 85}}, {"node": {"oid": "337473378333ab205b1679a008017f053005e00f", "messageHeadline": "Refactor ReduceMean/Sum Gradient without Shape Dependency. (#4261)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 87, "deletions": 100}}, {"node": {"oid": "deea945f80fe7bdb03fdc65e4382e7233b8f7df5", "messageHeadline": "Remove openmp and scipy from build pipelines (#4305)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 59, "deletions": 102}}, {"node": {"oid": "867ba846f7a6130319381070e66d18c5355ffa49", "messageHeadline": "Implement MinMax with SIMD (#4285)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 798, "deletions": 97}}, {"node": {"oid": "5ebdf76aa34afc205753c6e65dec85923e77a577", "messageHeadline": "Merged PR 4822584: change for dmlxp", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 8, "deletions": 6}}, {"node": {"oid": "4e39fda06ae89556e4ec1dde03fe67619e270bd8", "messageHeadline": "Fix version of torch and torchvision in install_deps.sh. (#4316)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "15cb4b3023dfd6e5439dcd8fc6151de4f3242a5c", "messageHeadline": "Fix session load state & run extra_postpasses only once (#4255)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 136, "deletions": 5}}, {"node": {"oid": "028b75e19f1d35c73f464f41abb2e539ae49ea09", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 494, "deletions": 800}}, {"node": {"oid": "d3c5cb6349b25e9fae927722179a3b5c90f770d1", "messageHeadline": "Use providers_available array from constants.h to avoid code duplicat\u2026", "author": {"name": "Prabhat", "email": "prabhatr@usc.edu", "user": {"login": "prabhat00155"}}, "additions": 1, "deletions": 31}}, {"node": {"oid": "737c22a911c6e178cc265cbd526b20e163563f7c", "messageHeadline": "Refactor Python packaging builds (#4283)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 493, "deletions": 769}}, {"node": {"oid": "3d6430959999ca690a3235ffed490e67d8fde67a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2741, "deletions": 2614}}, {"node": {"oid": "9e3b5c62fbb4db01b26ffd0947d2fe4c9ca0376c", "messageHeadline": "Use OpenMP-like synchronization patterns in Eigen thread pool (#4236)", "author": {"name": "Tim Harris", "email": "tiharr@microsoft.com", "user": {"login": "tlh20"}}, "additions": 917, "deletions": 653}}, {"node": {"oid": "57fabfba7a97ee8b7a8df162c791be2ade32846b", "messageHeadline": "Added GetAvailableProviders() to C API (#4247)", "author": {"name": "Prabhat", "email": "prabhatr@usc.edu", "user": {"login": "prabhat00155"}}, "additions": 129, "deletions": 7}}, {"node": {"oid": "175983c082526806fd372a2112918f3e07624deb", "messageHeadline": "Move memory info into IAllocator (#2850)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 284, "deletions": 329}}, {"node": {"oid": "064afa0f93509c7865684a963442a8275ed0d29e", "messageHeadline": "define dim_idx before use it (#4290)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "2204d39a06cb4f1e32d1be9a4c94837807fd7d8d", "messageHeadline": "Add build option to disable traditional ML ops from the binary. (#4272)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 376, "deletions": 243}}, {"node": {"oid": "3c633384c233f51ef7fd2382ea30a8754fdc6af7", "messageHeadline": "Fix TensorRT memory leaks (#4227)", "author": {"name": "alkoumpa", "email": "66834395+alkoumpa@users.noreply.github.com", "user": {"login": "alkoumpa"}}, "additions": 45, "deletions": 35}}, {"node": {"oid": "a541d28fb412c850df115382da35e670ec89742c", "messageHeadline": "Lazily get allocator when allocating an MLValue (#4276)", "author": {"name": "Derek Murray", "email": "Derek.Murray@microsoft.com", "user": null}, "additions": 5, "deletions": 2}}, {"node": {"oid": "a490beedf10d71c4a69eda4181e5dea5c374e0b6", "messageHeadline": "update tvm submodule (#4284)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e08181f74dc1e6cd0c13ba9b844d6be9a589d7a3", "messageHeadline": "Update Bert Notebooks for ORT 1.3.0 (#4274)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 570, "deletions": 1092}}, {"node": {"oid": "466511c1c3cc65015967ecc61f30e2ee5df38869", "messageHeadline": "Update gpt2 benchmark with position_ids and fp16 (#4275)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 408, "deletions": 249}}, {"node": {"oid": "efb99c24a75bc4d2b7482412c3581db9e5bcd6e4", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2182, "deletions": 781}}, {"node": {"oid": "0349479b19d7735956be3cfb0bc1c8014a214dbb", "messageHeadline": "Fix component governance and codesign validation errors (#4277)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 77, "deletions": 26}}, {"node": {"oid": "d5610e666ba96111470ba267d3b451df9901a035", "messageHeadline": "Support CUDA kernel for Einsum op (#4095)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1635, "deletions": 730}}, {"node": {"oid": "478b923e1990f25205565f33244e22ab12667360", "messageHeadline": "Expose ACL/ARMNN providers to Python (#4260)", "author": {"name": "goloskokovic", "email": "goloskokovic@gmail.com", "user": {"login": "goloskokovic"}}, "additions": 57, "deletions": 7}}, {"node": {"oid": "e505faa022e38e2b56bebcab1655fd8437327b40", "messageHeadline": "Fix two compiler warnings (#4263)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "5d773ee57bb0f0031c760475946bb4b0a6e73f59", "messageHeadline": "MLAS: add sgemv path for aarch64 builds (#4254)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 406, "deletions": 11}}, {"node": {"oid": "5da849b414711451b934c83ea8d55d898cceb7bd", "messageHeadline": "Fix detection of protobuf with onnxruntime_PREFER_SYSTEM_LIB on Linux\u2026", "author": {"name": "Chih-Hsuan Yen", "email": "yan12125@gmail.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "43deec2174164eddc6289344aee2608bc7a94b4b", "messageHeadline": "Temporarily remove dnnl from Linux CI build to unblock the whole team\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "1e1ba6cc4f59b20f0ae5e405039ebfcc5e92e02a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 668, "deletions": 657}}, {"node": {"oid": "1de79d315824f62d5ce4a4dc42686ac8eb730fc6", "messageHeadline": "Merged PR 4781392: Add 64-bit tensor types to DML EP", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 603, "deletions": 373}}, {"node": {"oid": "b41fcf1570b5ad771fe65f85b758f5db3bb0cd11", "messageHeadline": "Bugfix for shape inference and GetShape. (#4243)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 29, "deletions": 4}}, {"node": {"oid": "12367a6b11e96e70112026cb7975f7035c7ad0ed", "messageHeadline": "[C#] enable string-typed FixedBufferOnnxValue in input (#4178)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 46, "deletions": 14}}, {"node": {"oid": "189fb60ef97ba715cdbd4a35afbc7155992c2c7b", "messageHeadline": "Fix a bug and add code to profile memory (#4241)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 124, "deletions": 36}}, {"node": {"oid": "63bf587623f75cd4299f569bf24290a1628a934b", "messageHeadline": "Use azcopy to download test data (#4221)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 111, "deletions": 59}}, {"node": {"oid": "61fa5476d55d98f6fb66b5d7b076169073bdb2c8", "messageHeadline": "Update PyTorch Bert notebooks (#4239)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 358, "deletions": 544}}, {"node": {"oid": "f9afd990dcd66fd7cd451f5b084da3627ec10d09", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2005, "deletions": 447}}, {"node": {"oid": "7ccce4379e24c91afbfa8cee7a994e47599103fa", "messageHeadline": "Improve fast_divmod (#4224)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 36, "deletions": 106}}, {"node": {"oid": "825392c25b6c41f575be0dd531c7d2c5e04c76bf", "messageHeadline": "Fix ORT server CI build (#4165)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "5d28efd4348b42c3f1ee7337a0c0352f99e6836b", "messageHeadline": "opset12 code cleanup (#4242)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 8, "deletions": 23}}, {"node": {"oid": "e0334f177cb3cc5dc7aeabe8a5ef4773cd9ea902", "messageHeadline": "Opset12 upgrade for existing models used by perf/e2e pipelines (#4238)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 146, "deletions": 13}}, {"node": {"oid": "4486c66ed49fffbe2f76d7829e02b069874dda21", "messageHeadline": "enable conv transpose 3D (#4218)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 139, "deletions": 9}}, {"node": {"oid": "b08771f00ead970b0ce660e1ea06842ef9747786", "messageHeadline": "Add ONNX Training Post-Passes to Front-End - Cont (#4041)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 704, "deletions": 24}}, {"node": {"oid": "0b5bbb16b8a36a374fcdfea7cb6b22cf862343d8", "messageHeadline": "Benchmark With IO Binding (#4206)", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 175, "deletions": 84}}, {"node": {"oid": "b4b1c6440a5f609b04a9c424edb89cb45dd533ca", "messageHeadline": "Enable ORT with CUDA 11 toolkit (#4168)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 178, "deletions": 53}}, {"node": {"oid": "88a9cceb414449bdcb1cd4c05a38c0e0c0d5110f", "messageHeadline": "fix relative links in CONTRIBUTING.md (#4212)", "author": {"name": "Emad El-Haraty", "email": "elharaty@gmail.com", "user": {"login": "emad"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d0d31efd861896c50c5012b1205f81c293e5ef76", "messageHeadline": "fix transformer doc format (#4003)", "author": {"name": "Guoliang Hua", "email": "32868157+nbcsm@users.noreply.github.com", "user": {"login": "nbcsm"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ecc901717e2c758b296a21c168158b4b12bf7a15", "messageHeadline": "Use subset to release gradient tensors earlier (#4222)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "886befaba1100dc8d2aa183669c75f5549b6d71a", "messageHeadline": "Add BatchNorm and Concat to ACL EP (#4190)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 423, "deletions": 27}}, {"node": {"oid": "877862184ef21cebfc881118eb1500f7e13042f9", "messageHeadline": "Fix subgraph based reshape fusion (#4185)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 124, "deletions": 33}}, {"node": {"oid": "bf3c32166d26c4bafaafe63d5de5f5dfec6494a9", "messageHeadline": "fix optional input/outputs (#4229)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "5708c4feaf840ccfba95e3c087f87de75055800a", "messageHeadline": "Handle corner case in Resize op (#4183)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "7a96cfc8f5a3ecf93bfa9c6517c9ebaaec04f74a", "messageHeadline": "operator code cleanup (#4228)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 57, "deletions": 84}}, {"node": {"oid": "e4d94ee0e38a7eb48e376dce6ccd9611011f64f4", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 60, "deletions": 58}}, {"node": {"oid": "c55f6d76bec46bdef0155debf03e3401af1e4ba2", "messageHeadline": "[Vitis-AI EP] Fix to enable multi-output subgraphs inside Vitis-AI EP\u2026", "author": {"name": "jornt-xilinx", "email": "65705538+jornt-xilinx@users.noreply.github.com", "user": {"login": "jornt-xilinx"}}, "additions": 60, "deletions": 58}}, {"node": {"oid": "e8d4a3d01b04627fb1ab2305410be9ab8f13e5dd", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 729, "deletions": 157}}, {"node": {"oid": "de9da123cf1a13e2cee7e39a6cfd4d8c8d0793cc", "messageHeadline": "Enable static memory planning for pipeline. (#4204)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 113, "deletions": 14}}, {"node": {"oid": "b377266eb339059044a39fdd041da59554a66186", "messageHeadline": "Fix Mac build linker warnings (#4155)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "91a41298cc88719a58cc2898c0d9ee54c8575e1c", "messageHeadline": "Fix ORT build when onnxruntime_PYBIND_EXPORT_OPSCHEMA is enabled (#3954)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 16, "deletions": 12}}, {"node": {"oid": "155e22d1ab7dd8afeb19b8cab86df4d53228d70c", "messageHeadline": "MLAS: fuse float output into quantized GEMM (#4215)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 593, "deletions": 131}}, {"node": {"oid": "2e3607c7cd4a25c66b189b2d50f8c2399c4824aa", "messageHeadline": "Remove hardcoded desktop lib (#4193)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "01edd6137233c0955f18b561db3e7947080cc04a", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 2830, "deletions": 575}}, {"node": {"oid": "f74861841e96930d946c4b2d784d351fdd08af6b", "messageHeadline": "Fix dangling pointer to local string variable in onnxruntime_pybind_s\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 8}}, {"node": {"oid": "6b4f652017b16618d0950df074e21ef201e38ccb", "messageHeadline": "Clean up status checks in gradient_graph_builder_test.cc.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "7096e6f5efd450e8c168737f74bd1946fdc43e6a", "messageHeadline": "Reduce severity of GraphAugmenter logging statement.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6f4320fb856acb849c22dfccbebf8455305986b1", "messageHeadline": "Fix the python package name issue (#4207)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 26, "deletions": 14}}, {"node": {"oid": "87d68d8531e428e266e9381760269b03bdc36d5c", "messageHeadline": "matmul integer fusion (#4195)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 852, "deletions": 99}}, {"node": {"oid": "2605faef88714b9e9d5ad0eac7f17c387db4cc1c", "messageHeadline": "Add past state support in Attention Op for GPT-2 (#4107)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1727, "deletions": 398}}, {"node": {"oid": "e6ccb1ac282ba77ead42651b0d1de0ebb49026b0", "messageHeadline": "GatherNDGrad for CPU (#4123)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 210, "deletions": 45}}, {"node": {"oid": "56a2dfd7d5bc3cf4fba25ecbe0dc6465f3e11ffc", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 195, "deletions": 65}}, {"node": {"oid": "65a682354b7d14d9a6f38ccfe1bcc71c5e1e5294", "messageHeadline": "enable pipeline to run with mixed precision (#4113)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 129, "deletions": 60}}, {"node": {"oid": "8f8d899bf2f8b7ebd63acb82f4c517cfad772571", "messageHeadline": "Enable code sign in c api pipeline and python pipeline", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 47, "deletions": 2}}, {"node": {"oid": "73bc6be5d15d466297a73ec546933db60c9d62cd", "messageHeadline": "build: split nodejs binding build and test to avoid timeout issue (#4\u2026", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 17, "deletions": 3}}, {"node": {"oid": "117b2e7743aea93fae5ac503195b230ac60d2b45", "messageHeadline": "Fix GPU memory leak on TensorRT (#4172)", "author": {"name": "Matthew Hill", "email": "matthill@gmail.com", "user": {"login": "matthill"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "e323411e316f86e6d6ed55621b524100b23dc236", "messageHeadline": "Merged PR 4790197: Remove hardcoded desktop lib", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "eae841464caa9d1fe4b5f55a57a80b310a96867c", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 478, "deletions": 433}}, {"node": {"oid": "af0750ba1ba08ecc27ef76059c3cd2d8cd191539", "messageHeadline": "Java GPu artifact naming (#4179)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 45, "deletions": 25}}, {"node": {"oid": "e8ed14bcb37611a472ec7cfdf8fe9aca87d67d1f", "messageHeadline": "disable MEMLEAK CHECKER for openvino", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c296884fc34965b23b453f1e09656835ffd0496e", "messageHeadline": "bump up ORT version to 1.3.1 (#4181)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 13, "deletions": 7}}, {"node": {"oid": "c0bdbc0b397267f46db1055cf4336a1f0ec531ff", "messageHeadline": "Enable telemetry for the C API and python pipeline (#4174)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 12}}, {"node": {"oid": "35d9f396c460b284022b2faad392b9523f3c3732", "messageHeadline": "MLAS: refactor quantized GEMM loops (#4182)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 405, "deletions": 385}}, {"node": {"oid": "9d65ce53bc09da1c870d04ef7a1eb715d95ca6fb", "messageHeadline": "move back to toolset 14.16 to possibly work around nvcc bug (#4180)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a7366d82af669e995f94621ddf9ac5c9d4788031", "messageHeadline": "Disable nuphar large model test (#4173)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "a47158e3ee58ba5a38748531757b76dba98e7ba0", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 83, "deletions": 4}}, {"node": {"oid": "9eba9fba7ce4b0ce7e982313116a766a3ae17c2d", "messageHeadline": "Fix for BiasGelu fusion optimizer (#4160)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 69, "deletions": 4}}, {"node": {"oid": "2b3ce1b0909277d3f64c53dedb67579d18d27b39", "messageHeadline": "add script to support update nodejs binding version (#4164)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "559d2c0c9c19affd98ccb7183db7417852a5fa16", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into DmlDev", "author": {"name": "ISS Build Account", "email": "issblder@microsoft.com", "user": null}, "additions": 156, "deletions": 50}}, {"node": {"oid": "a9862b00f30d0d5a1993ce6cc49363036fa80b57", "messageHeadline": "Merged PR 4779170: Merge ORT master", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 8613, "deletions": 2296}}, {"node": {"oid": "4377ff4a1a1d0ec65a811669abcf5f0508007469", "messageHeadline": "Enable .NET Core 2.0 and .NET Framework 4.6.1 in Microsoft.AI.Machine\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 132, "deletions": 19}}, {"node": {"oid": "28d12dc4f073a3e8e2e5a70b86b5177f6c6d9cbc", "messageHeadline": "Try to avoid std::move in return whilst keeping CentOS build happy. (\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 7, "deletions": 5}}, {"node": {"oid": "541eafb41a4f141d33b34137b834b929e4f92469", "messageHeadline": "Fixed the link to model test documenation (#4011)", "author": {"name": "oak-tree", "email": "k.o.alon@gmail.com", "user": {"login": "oak-tree"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2ab3a19728c551635417a73d21c95dda50bbe504", "messageHeadline": "Enlarge the read buffer size in C#/Java test code (#4150)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 16, "deletions": 25}}, {"node": {"oid": "045ce7d13b5bea2698d87a12fd383c8191daa1a5", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/ticastro/upd\u2026", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 8613, "deletions": 2296}}, {"node": {"oid": "8eb6a539bd32f56f2c9b9c6d510cbeefe11846bf", "messageHeadline": "Hardcode WinML tests umbrella lib (#4161)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7f5339505e146bfd303b629cd8f7d052ac72b461", "messageHeadline": "Discover trainable parameters using reverse DFS from loss node (#4116)", "author": {"name": "suffiank", "email": "suffiankh@gmail.com", "user": {"login": "suffiank"}}, "additions": 119, "deletions": 44}}, {"node": {"oid": "842be1535d1ac06edef134e7f52eb838a2c1692c", "messageHeadline": "[Node.js binding] add linux and mac package (#4157)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 193, "deletions": 57}}, {"node": {"oid": "653417ae4bdd6b166006e2aa995be7868ac04447", "messageHeadline": "Fix scaler->scalar typo. (#4142)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 2, "deletions": 1}}, {"node": {"oid": "6bbd18efd0f5b04ada6bd95bfb92ef3a8456d72b", "messageHeadline": "Hardcode WinML umbrella lib to windowsapp.lib (#4133)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "ee3532097477adf4e53aec270a782dd228d3625d", "messageHeadline": "The fixings for python scripts in ONNXRuntime (#4135)", "author": {"name": "Wenbing Li", "email": "10278425+wenbingl@users.noreply.github.com", "user": {"login": "wenbingl"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "3390431d806881f927e8b3f27fed59fc4413bdb5", "messageHeadline": "Update MCR image table (#4137)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 9, "deletions": 8}}, {"node": {"oid": "5a5f44eed70dbbdbd82584712c2ba50ef28581cc", "messageHeadline": "Add softmax to the mnist example (#4149)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 1}}, {"node": {"oid": "4e1dac67cd13b9cc2ea0c936a6c3d05f4ad27b2b", "messageHeadline": "Address memory leak and improve memory handling (#4124)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 200, "deletions": 102}}, {"node": {"oid": "b8db8076cb0a55acf1621c152c09e88fc19ce975", "messageHeadline": "Fix MKLML Tests Run (#4144)", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 16, "deletions": 12}}, {"node": {"oid": "7c8e1580a13ce333e47a41146bccfc90b3a70db5", "messageHeadline": "Add check of graph output in Bert Fusions (#4126)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 104, "deletions": 76}}, {"node": {"oid": "ffed43e9b8aa37b9fa8cc08a0617a94be53a2df1", "messageHeadline": "handle loss and name marching wrappers (#4066)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 240, "deletions": 81}}, {"node": {"oid": "2aab20b4ea7ce69ce804cc1bdbc6bc951233c721", "messageHeadline": "[Node.js binding] upgrade node-addon-api to 3.0 (#4148)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 144, "deletions": 127}}, {"node": {"oid": "ea368f69db20bc2fb93fbebabd7dbef645c1407e", "messageHeadline": "Add Swift/macOS sample, a port of the Windows MNist sample", "author": {"name": "Miguel de Icaza", "email": "miguel@gnome.org", "user": {"login": "migueldeicaza"}}, "additions": 1664, "deletions": 0}}, {"node": {"oid": "2e58097f8f12f9bb3fe9f1e593793fb718ebf454", "messageHeadline": "fix build: pipeline Node.js version to 12.16.3 (#4145)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "1e5307d4588b8399374fd5edc8cb1e48d7831a03", "messageHeadline": "Bug fix for parameter names of models not using wrapper (#4061)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 59, "deletions": 6}}, {"node": {"oid": "9790e194249b4e1a3d4a8d5ab415a55153174b5e", "messageHeadline": "Handle mem pattern allocation failure better. Make BFCArena behavior \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 134, "deletions": 58}}, {"node": {"oid": "81101c9efd15a69342dec94362cf7c29a4e41be9", "messageHeadline": "Fix DropoutGrad op (#4052)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 19, "deletions": 2}}, {"node": {"oid": "6199ef1375c7133c04e64d253ae7c5149455dd67", "messageHeadline": "Change group id to com.microsoft.onnxruntime per requirements.", "author": {"name": "Dmitri Smirnov", "email": "dmitrism@microsoft.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "16cef90e29f2dac3a3612994eb33cb8de1827671", "messageHeadline": "General enhancements/cleanups to test exes (#4109)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 357, "deletions": 275}}, {"node": {"oid": "197da135eb4d0e4db74789298a9726c8924711e2", "messageHeadline": "Implement quantized Attention on cpu (#4111)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1059, "deletions": 273}}, {"node": {"oid": "62b44527e5d25ea170c9433654f189142f95cff6", "messageHeadline": "Add ArmNN Execution Provider (#3714)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 1530, "deletions": 6}}, {"node": {"oid": "62af8da3f6231f6b85744f134f4ad8bf3894e452", "messageHeadline": "Use OrtMutex and OrtCondVar everywhere instead of std::mutex/std::con\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 38, "deletions": 37}}, {"node": {"oid": "905c53562601ca4940bed2e445c7fcc6445f7687", "messageHeadline": "still need to make the test stable. Lower the acc number a bit to mak\u2026", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d63b90538eb97ac281f7a28a809759846901e5d0", "messageHeadline": "Symbolic shape inference exit on models without onnx opset used (#4090)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 11, "deletions": 3}}, {"node": {"oid": "6f8a4f4cadc1affdcf0bd4abbb192c603669ddcd", "messageHeadline": "Fix Nuphar test failure", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 2, "deletions": 3}}, {"node": {"oid": "32d8a76f2f203fec9b8e535c5199fd7beb18822e", "messageHeadline": "Fix Nuphar build in gcc 7 (Ubuntu 18.04)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 14, "deletions": 14}}, {"node": {"oid": "f18a99b2452fe364c5fc7993c95d3d719bc14244", "messageHeadline": "Exclude non-trainable torch buffers from trainable weights (#4099)", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 36, "deletions": 2}}, {"node": {"oid": "e5cec7237d30e581902d20a667fea318fb40ca63", "messageHeadline": "Clarify telemetry collection (#4102)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "baa0697982cdda2fbfa107127efbdb4947143e4b", "messageHeadline": "[OpenVINO-EP] Add missing dependency libs in Dockerfile (#4064)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 14, "deletions": 8}}, {"node": {"oid": "647a8865871d0860bfea9564567dc0ae183a96f4", "messageHeadline": "[Nodejs binding] create a new pipeline to generate signed binaries (#\u2026", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 96, "deletions": 26}}, {"node": {"oid": "3f7b97a63da00cba3bb46bb9952b4baa796b3cd0", "messageHeadline": "MLAS: more code cleanup (#4101)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 424, "deletions": 288}}, {"node": {"oid": "08e5f89b37fef05343ab7e823259f4d4f571c097", "messageHeadline": "Fix the nuget gpu pipeline (#4106)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "afca0d15ee20a13895ec36ade2c9c9582e818353", "messageHeadline": "Create Java publishing pipeline (#3944)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 860, "deletions": 12}}, {"node": {"oid": "51d78bc5e607ce2af29ecd27ca145a9efeec94a4", "messageHeadline": "Fix DML EP doc link to C API (#4105)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6c1b2f33b74ad48c3cb08d4ba1f38e1897659c8e", "messageHeadline": "Fix crash reported in #4070. (#4091)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 93, "deletions": 9}}, {"node": {"oid": "8813d205cc57fe27887cd0bd010aa57a39aca668", "messageHeadline": "Update GPT2 Model Benchmark Script to Support IO Binding (#4088)", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 126, "deletions": 19}}, {"node": {"oid": "ba74914c5a1229459f995c66f937972bca3193bc", "messageHeadline": "Remove evaluation output from training e2e test baseline data. (#4092)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "3eaec57c3812b2e667f50e77f64a03b3c981e7b0", "messageHeadline": "Fix the daily pipeline failures (#4084)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 79, "deletions": 188}}, {"node": {"oid": "f54518bae999a9fbcd2c1f97e18ad01436d4be0d", "messageHeadline": "Actually switch the spdlog submodule to the master branch. (#4100)", "author": {"name": "Derek Murray", "email": "derek.murray@gmail.com", "user": {"login": "mrry"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "72d508b7a088ea942e1c8d98a70a665763b900de", "messageHeadline": "New perf metric - e2e throughput (#4085)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 222, "deletions": 220}}, {"node": {"oid": "70d91a85505ae42a14969c8eb57b5509abeb7c07", "messageHeadline": "re-enable graph optimizations during build phase (#4044)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 5, "deletions": 6}}, {"node": {"oid": "ff16ca54e1c0d08bc1c6ebaabd4ce6cf2425f738", "messageHeadline": "Fix the flake8 warning in generate_nuspec_for_native_nuget.py (#4089)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a715d55bcce47f251537577fbfbf5a90bb4fe4fd", "messageHeadline": "Training Python package fixes (#4063)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 40, "deletions": 37}}, {"node": {"oid": "9d748afff19e9604a00632d66b97159b917dabb2", "messageHeadline": "Set spdlog submodule branch to \"master\" explicitly. (#4087)", "author": {"name": "Derek Murray", "email": "derek.murray@gmail.com", "user": {"login": "mrry"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "1d441f89acc7c4d17ecda1fb41ca0a2310ef2a57", "messageHeadline": "Re-enable PEP8 check in Win CI build  (#4075)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 10, "deletions": 7}}, {"node": {"oid": "b85805ed013b0a9d5f9a750bf3ef4debb31ff401", "messageHeadline": "Handle edge case with implicit input and multiple levels of subgraphs\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 118, "deletions": 3}}, {"node": {"oid": "c331d8cffc77aa6aebcce8581138813eda9611cb", "messageHeadline": "WinML custom operator header is missing from nuget package. (#4083)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "6c7eaff6769165efc35b4b43e720c385849cab10", "messageHeadline": "fixed typo in readme (#4076)", "author": {"name": "Linnea May", "email": "linneamay98@gmail.com", "user": {"login": "linnealovespie"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6404aba5ae962604477e27833069279e1e0ccde7", "messageHeadline": "Orttraining rc1 master merge (#4080)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 72, "deletions": 35}}, {"node": {"oid": "e951b29a0bd1168d7233bc04d0c3498fa017241c", "messageHeadline": "Fix a macro and memory regression (#4068)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 139, "deletions": 80}}, {"node": {"oid": "38d76cc9046205b81f2b7af0698be7251f9f145d", "messageHeadline": "Clean up training E2E test (#4078)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 62, "deletions": 82}}, {"node": {"oid": "dd43623da2041de15536c44c9d1db62f5f40a2b2", "messageHeadline": "Remove ONNX from requirements.txt (#4073)", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "348ed698ec70c8e02c375afdf4f6c1cffb3ad887", "messageHeadline": "Add more symbolic compute support in symbolic shape inference (#4057)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 15, "deletions": 8}}, {"node": {"oid": "2a96be83f629bfbf2aa5c3937e9c6e86e050bdad", "messageHeadline": "skottmckay/bugfix/SubgraphInput (#4004)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 138, "deletions": 22}}, {"node": {"oid": "c55634d2e6f5680a2c759bf6a6dfe9827dc8df93", "messageHeadline": "Fix initial value of loop variable in RNN op (#4055)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 34, "deletions": 3}}, {"node": {"oid": "6d03470587412f63dbb343be3a130e178a2f189b", "messageHeadline": "Add e2e measurement for training (#4049)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 138, "deletions": 101}}, {"node": {"oid": "847bece23d62a0b5f96aab52dd8a09b01ce5d84e", "messageHeadline": "Merged PR 4735426: Merge latest github master enabling disabled tests", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 4098, "deletions": 6485}}, {"node": {"oid": "5bd82e913d37b458bb555719e59fd123c6fda711", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/jamather/mer\u2026", "author": {"name": "jake mathern", "email": "j.mathern@digipen.edu", "user": {"login": "Jamather"}}, "additions": 4098, "deletions": 6485}}, {"node": {"oid": "26be762b35e69b6edf9ac4086641fa569f27f6f9", "messageHeadline": "Make CPU QuantizeLinear support optional zero point (#4065)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 18, "deletions": 11}}, {"node": {"oid": "60fa4b1f905f55c9734bcd9bf3572f7bea738102", "messageHeadline": "Update benchmark of gpt2 model with past state (#4043)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 163, "deletions": 117}}, {"node": {"oid": "ed0a8e5b5c4be40c347ab35d055e3e258cbf9df0", "messageHeadline": "Enable disabled tests and add fixed model (#4059)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 35, "deletions": 38}}, {"node": {"oid": "279f9aa8659e10132b68755ce2b0fb8f49ab543d", "messageHeadline": "Update WinRT_API.md to reflect 1.3 release (#4074)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "c94d9685b644da6ebb52c92d7e85dc24fce84766", "messageHeadline": "Fix a problem in StacktraceTests::BasicTests (#4069)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "a859dc422c17c71bf147e7dde1648e1c6bcfbaae", "messageHeadline": "Delete google::protobuf::io::FileInputStream class from our source co\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 283}}, {"node": {"oid": "1e82ecfd5c4542548256eef805927269a7b46687", "messageHeadline": "Fix link in readme (#4058)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7f750b65cef3255891e20d412e16b9f787a4c6a7", "messageHeadline": "support model > 2GB in transformer optimizer (#4038)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1786, "deletions": 5307}}, {"node": {"oid": "9f7d2454463e19f585f3acd75e2ca5b128a88a55", "messageHeadline": "Add noexcept to various OrtCallback utility class methods to fix warn\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "23c313cb73e838f6c35052a7fd05ca8131dbfbbd", "messageHeadline": "fix crash in dequantizelinear/quantizelinear for optional zero point \u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 86, "deletions": 88}}, {"node": {"oid": "6665d5e2bc3dc8cbfcd85a3d3fadb91aebddfe86", "messageHeadline": "Liqun/a transformer example (#3845)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 582, "deletions": 3}}, {"node": {"oid": "a983509ed391cfd94f571b6f72cba0211cb4fbee", "messageHeadline": "Pad: Add support for all datatypes in opset-11 spec (#4021)", "author": {"name": "Matthieu Darbois", "email": "mayeut@users.noreply.github.com", "user": {"login": "mayeut"}}, "additions": 452, "deletions": 409}}, {"node": {"oid": "930c6a59dac64357ec15015724c08a4fb48b167f", "messageHeadline": "Allow optional cast in embed layer norm be optional. (#4040)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 50, "deletions": 8}}, {"node": {"oid": "b3ec8035eef4735cd919d11f5f970ca8a1fcd87b", "messageHeadline": "[Node.js binding] add build flag for node.js binding (#3948)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 184, "deletions": 50}}, {"node": {"oid": "ee6371d0a8165818ba6fbf91e8b59114c2afc4a3", "messageHeadline": "Clean up CUDAExecutionProvider's associated PerThreadContexts on dest\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 115, "deletions": 82}}, {"node": {"oid": "60aada3862edd19487b9c5e2d36197368e4287e1", "messageHeadline": "Merged PR 4716166: Merge latest ORT Github master into dml dev", "author": {"name": "Ryan Lai", "email": "Ryan.Lai@microsoft.com", "user": null}, "additions": 15543, "deletions": 4454}}, {"node": {"oid": "633008b5ef1ef2fca7d16aef667b3d867a1b558d", "messageHeadline": "Add pipeline online partition logic for pipeline (#3996)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 612, "deletions": 77}}, {"node": {"oid": "c2c45c671599fa5e13ab35be7e1f13c2dd6aff91", "messageHeadline": "Merge https://github.com/microsoft/onnxruntime into user/rylai/fix_gi\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 4156, "deletions": 1567}}, {"node": {"oid": "0d8abc1a99d59143eecaff4767c800882b367265", "messageHeadline": "MLAS: qgemm refactoring (#4030)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 738, "deletions": 1386}}, {"node": {"oid": "abcd1576c900889917939abf87bd7ad8b08ea0a9", "messageHeadline": "Add Linux bash and Windows batch scripts for running transformers ben\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 303, "deletions": 0}}, {"node": {"oid": "212efb6cde0d8abd0b9dcb0a8ef9ed7154c1f81b", "messageHeadline": "Match New Pattern for Reshape Fusion (#3931)", "author": {"name": "Cecilia Liu", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 279, "deletions": 48}}, {"node": {"oid": "7759136610cd8415264892ad83acdf9d8ed7e87a", "messageHeadline": "Add amd migraphx execution provider to onnx runtime (#2929)", "author": {"name": "Paul Fultz II", "email": "pfultz2@yahoo.com", "user": {"login": "pfultz2"}}, "additions": 1957, "deletions": 18}}, {"node": {"oid": "9d0534c0ebce73265cbdcb7863c474c26daf98c9", "messageHeadline": "Optimize OneHot CUDA Kernel (#4012)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 371, "deletions": 5}}, {"node": {"oid": "0a6d9dd301cb0a7124305108546d70f6eee5f244", "messageHeadline": "Remove Openmp from the GPU docker files", "author": {"name": "Changming Sun", "email": "me@sunchangming.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "30efe65e95a8b0b104c8964f0f31a1d2b1d8f3d2", "messageHeadline": "Add use_openmp back to the docker files", "author": {"name": "Changming Sun", "email": "me@sunchangming.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "bd8993cb15c0ca3debfe9f54ed63c30aed972c82", "messageHeadline": "remove --use_openmp in build.sh", "author": {"name": "Wenhao Hu", "email": "fumihwh@gmail.com", "user": {"login": "fumihwh"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "faf65e960f0e9ba0af3857c90d2c1aad122b8e8f", "messageHeadline": "Refactor delayloading (#4019)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 33, "deletions": 23}}, {"node": {"oid": "24eda3df3318d86f310918ad9021295d4a5c7898", "messageHeadline": "Create Utils for Adding Range and Marker (#4013)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 461, "deletions": 9}}, {"node": {"oid": "aafe988a1166732bc4c437f774ad4827aacbf043", "messageHeadline": "Temporarily disable windows static analysis CI job", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 67}}, {"node": {"oid": "7c831183647e919f2b778a93e9d98171d3f4871d", "messageHeadline": "Enlarge protobuf read buffer size", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "0868a100d553d9a102cbe64e4d51eb6c149a6b42", "messageHeadline": "Merge https://github.com/microsoft/onnxruntime into user/rylai/fix_gi\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 11394, "deletions": 2894}}, {"node": {"oid": "eb3aaa70d6ca40139e2407464898b59595807581", "messageHeadline": "Fix compiler warning for openvino  (#4010)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "59af3ea278238b9f92cb9793533a61490a019797", "messageHeadline": "Add missing D3D12 resource barriers and fences to Winml (#3941)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "ce4d05862ad3c297fe4b415a36eb559a7214dfba", "messageHeadline": "add bm_fish_720 to collateral for scenario 22 test (#3998)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "357bffe47cd2f665b118f3bb170e431e3b6bf0e2", "messageHeadline": "Fix deprecated CentOS link for Linux CI pipeline (#4000)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0a5395bb78810d0beec2c4b0f9094998580f45aa", "messageHeadline": "Remove 'model_.' prefix from onnx model initializers in training (#3881)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 179, "deletions": 229}}, {"node": {"oid": "08763e80e0be154da1700ec94b4d489e86b1efe7", "messageHeadline": "Fix permission denied while creating directory in azure pipelines (#4\u2026", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 7, "deletions": 4}}, {"node": {"oid": "dbd5aab6d2a0dc8858b6be4c1dd08212e30f851a", "messageHeadline": "Update OnnxRuntime.java for OS X environment. (#3985)", "author": {"name": "jji2019", "email": "49252772+jji2019@users.noreply.github.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "34cdd9ab0ef838f009e945efa16b540bd9b45a9d", "messageHeadline": "Merged PR 4687622: ORT DML EP Pads-11 does not support float16 (falls\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 32, "deletions": 3}}, {"node": {"oid": "989fe2498f4dd0beaf1b4324bf74a39abaf86a29", "messageHeadline": "Change training perf test build to use \"docker\" instead of \"sudo dock\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 36, "deletions": 29}}, {"node": {"oid": "354e5712776643b77accb948a6ce46faa01d82a9", "messageHeadline": "Miscounted the number of characters in package version of DirectML nu\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4e84df25e792dde576fded532251a5245570bf99", "messageHeadline": "Merged PR 4700287: Fix 5d concat operator helper", "author": {"name": "Jake Mathern", "email": "jamather@microsoft.com", "user": {"login": "Jamather"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2adb0f8d1d28a268ba33994958fee3aacb5f6d33", "messageHeadline": "Fix 5d concat operator helper", "author": {"name": "jake mathern", "email": "j.mathern@digipen.edu", "user": {"login": "Jamather"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "475e7e43e6f6e4573daa92f8a82df4ebe561ff09", "messageHeadline": "Older flake8 versions report false positives and don't handle the sam\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 51, "deletions": 31}}, {"node": {"oid": "fb4efafc8e2e5cf4361b11538aa80bc12eed439f", "messageHeadline": "GPT-2 training perf scripts (#3974)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 173, "deletions": 70}}, {"node": {"oid": "36bcb28238f3555d080f998d2f671f835f24d066", "messageHeadline": "Add NNAPI in the exclude list (#3921)", "author": {"name": "Charles Lien", "email": "lienching@users.noreply.github.com", "user": {"login": "lienching"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "64b5f7edf6e5464d94c61cf6f3f16442b441e583", "messageHeadline": "Initial release of Vitis-AI Execution Provider (#3771)", "author": {"name": "edelaye", "email": "edelaye@yahoo.com", "user": {"login": "edelaye"}}, "additions": 953, "deletions": 6}}, {"node": {"oid": "c42867c01698d400e9b9ff1c11d7f7ec13250195", "messageHeadline": "Update data_frame_tool to latest (#3919)", "author": {"name": "Gani Nazirov", "email": "ganinz@hotmail.com", "user": {"login": "ganik"}}, "additions": 59, "deletions": 12}}, {"node": {"oid": "672c42b396b8feae5a424d40bc0f9fa2315605d1", "messageHeadline": "Fix std::_Exit workaround for Dnnl crash at unload (#3978)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 29, "deletions": 19}}, {"node": {"oid": "b8a255e1b5509c0fa93f0aae9708dfc2dab13dd9", "messageHeadline": "Doc Updates for Build  (#3976)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 422, "deletions": 512}}, {"node": {"oid": "1168f4e85a90815da9b065df05e056c77c25c8d6", "messageHeadline": "Support session EndProfiling() in the CSharp API  (#3934)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 85, "deletions": 5}}, {"node": {"oid": "b253f0b0f677ae55f06aea2f6e65e12486686b82", "messageHeadline": "Do not fuse skiplayernorm if any output of add in graph output (#3981)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 27, "deletions": 9}}, {"node": {"oid": "a75a83b41a9c64a76dbf899b501a7087a60221ab", "messageHeadline": "Minor android build fix (#3980)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 10, "deletions": 3}}, {"node": {"oid": "2fa2019daf0588757e77d67305fdcb95cc0d54a7", "messageHeadline": "Run docker commands with sudo (#3979)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 56, "deletions": 56}}, {"node": {"oid": "29c39f867c6e1a5fe340883764d040d3f45a1dec", "messageHeadline": "enable tensor core for rnn (#3937)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 22, "deletions": 6}}, {"node": {"oid": "c6a94f95cfd0bde49213ea90800d29ee106fbc49", "messageHeadline": "Update Android instructions (#3971)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 174, "deletions": 14}}, {"node": {"oid": "024b92a970d9ca8a23bd3a04605244ded228459d", "messageHeadline": "Use path relative to script location to refer to symbolic_opset10.py \u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "9d2d1eb6f65c1354e5557364b9d218357ae31705", "messageHeadline": "[java] Adds a CUDA test (#3956)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 43, "deletions": 5}}, {"node": {"oid": "1a183784a8d79c8b7fe19184f43e82945ae3084c", "messageHeadline": "Fix C# layer in the way it handles sequences (#3965)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 77, "deletions": 2}}, {"node": {"oid": "e259a13f8ec1804c93a837ed6fcc2522d8a4978a", "messageHeadline": "Initial training Python packaging pipeline (#3767)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 627, "deletions": 31}}, {"node": {"oid": "e55f24364af38b1fa81eff5f87cc768133b774f8", "messageHeadline": "Disable LTO on Windows training CPU build (#3960)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "44731e88bbe36432db0c0f7b25bd1492a0e92c70", "messageHeadline": "Add comments for zero valued normalization factor in SoftmaxCrossEntr\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "fd8ea4e4661a3333b36911330b057dba10ada9db", "messageHeadline": "Improve handling of symbolic dimensions in the onnxruntime_test.py sc\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 20, "deletions": 2}}, {"node": {"oid": "523d70f6675c0558e3e1064d0d74feaad593875e", "messageHeadline": "Improve Transformer Benchmark for FP16 (#3970)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 78, "deletions": 57}}, {"node": {"oid": "0d11649bb304a801e6e3743d72cf61c11db596d2", "messageHeadline": "Address comments from #3823 and polish code (#3964)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 107, "deletions": 72}}, {"node": {"oid": "4ff73d00b0670668027827f792d5a0a26664bb10", "messageHeadline": "Fix python pkg permission issue (#3957)", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 6, "deletions": 3}}, {"node": {"oid": "07e9a4c16404dac4a9bd57f671af764294a35a82", "messageHeadline": "Update benchmark to reflect those used in our latest results (#3967)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 270, "deletions": 166}}, {"node": {"oid": "56700bec835d47e56a005d5e8ef708f6bd238485", "messageHeadline": "Add example of python code to readme of transformers tools (#3966)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 28, "deletions": 5}}, {"node": {"oid": "769c11f217035d229d87ae83c0a285a3ea9df1ae", "messageHeadline": "Update doc for transformers tools (#3963)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 77, "deletions": 81}}, {"node": {"oid": "a296b167199783770d9a233ceb31f7e4c8d8fe40", "messageHeadline": "Prevent divide by zero in CUDA implementation of SoftmaxCrossEntropyL\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "132ce3a56148b93cc00bfdbdfbe695ddb74936eb", "messageHeadline": "Fixes for quantizing a BERT from HuggingFace (#3939)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 36, "deletions": 30}}, {"node": {"oid": "33208c9f6bad07512a67bc953ef9886953736fa3", "messageHeadline": "Modify Pipeline Facilities to Fix PipeDream Deadlock (#3823)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1202, "deletions": 415}}, {"node": {"oid": "999554cc532e38aa92609dd0d0a438f40bc28211", "messageHeadline": "CGManifest - add training entries and generate entries for submodules\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 759, "deletions": 185}}, {"node": {"oid": "bc441b7e5c3a7544e91854e5be582dab270a9a3c", "messageHeadline": "Add cpu/mem usage for perf metrics (#3947)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 36, "deletions": 29}}, {"node": {"oid": "be003dbab7e9117acec04cc118bbeb4e67ebd8cb", "messageHeadline": "Fix ACL build break (#3952)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "47ae9691fdb817e9ac85773526f7a6739e94d962", "messageHeadline": "Fix ordering of APIs. (#3951)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "9ef376880b4ebec138ab9e25cb95df3912d80ef4", "messageHeadline": "Add test for If node with conditional branches only containing Consta\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 74, "deletions": 0}}, {"node": {"oid": "38467f8c9a8f34c36ea4e48aaddb9298b9895f1d", "messageHeadline": "DirectML Nuget package has different time stamp than Native and Manag\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 15, "deletions": 2}}, {"node": {"oid": "e6da5946d1c7ccaef227f59d9357e74f9d649c1b", "messageHeadline": "Update DML Nuget version and DML EP Doc (#3945)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 11, "deletions": 10}}, {"node": {"oid": "782c6c24b2d52b2fcf4fcbb129567f30f9649a04", "messageHeadline": "Rename bert to transformers (#3946)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3844, "deletions": 87}}, {"node": {"oid": "3c4f3d01cd2a79a3fb7f93edcb6d4b58fe4993ac", "messageHeadline": "Implement QLinearLeakyRelu (#3648)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 236, "deletions": 0}}, {"node": {"oid": "5e0928a777745a1536332973418dbbb7ccbf9d52", "messageHeadline": "Enable running PEP8 on python scripts using flake8 (#3928)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 639, "deletions": 524}}, {"node": {"oid": "39814319b2d3051d7cf1bf7bcde6f71c296ad90c", "messageHeadline": "[Node.js binding] fix linux build (#3927)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "50f798dad6681b5f84ece1a97b4a90504aa330f0", "messageHeadline": "support non-zero zero point for matmulinteger u8s8 (#3883)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 139, "deletions": 160}}, {"node": {"oid": "f57515aae40044fd9c9f2b7acb4c7a7a307313a8", "messageHeadline": "Update DML nuget to 2.1.0", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "e9a053ff7fd5adbe41c289784908b8079a9093a4", "messageHeadline": "Fix version number in doc", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "756767ea78383991701c0299e4e9fed39a79c956", "messageHeadline": "Update DML EP Doc for ORT 1.3", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 8, "deletions": 7}}, {"node": {"oid": "9c989c8dd687f030239c48cb56ec6952eba542a4", "messageHeadline": "Update build doc for cross-compiling (#3672)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 178, "deletions": 28}}, {"node": {"oid": "cab21223b3eb5a531bc7841dd99c8151f20260de", "messageHeadline": "Updated TPN for OpenMPI and cleanup (#3932)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 113, "deletions": 4}}, {"node": {"oid": "cba8bdc79001f4dfbd2e92c81ac75a6b95d71818", "messageHeadline": "Make some compile change for Android NNAPI provider using DNNLibrary \u2026", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 27, "deletions": 30}}, {"node": {"oid": "84c108a85e55093b38f5be09f26fb3ad56ff2ac3", "messageHeadline": "link to folder instead of READMEs inside folder (#3938)", "author": {"name": "Prasanth Pulavarthi", "email": "prasantp@microsoft.com", "user": {"login": "prasanthpul"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "48f69cfbb8effe30ece62f03504a99a92d5759a7", "messageHeadline": "Fix DirectML nuget creation in Nuget pipeline (#3929)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f380460a9efbc7513d59a9ea23326da131b37357", "messageHeadline": "Update the build steps to support ORT on Jetson (#3869)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 28, "deletions": 2}}, {"node": {"oid": "d80e15fb1112821cc7e204d83aa47c3e1579f7eb", "messageHeadline": "Fix DmlCopyTensor test (#3923) (#3925)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 11, "deletions": 20}}, {"node": {"oid": "93eb9bcfdec2e3204d6a420939a112ee10e4550a", "messageHeadline": "Add yaml/perf scripts for new perf test pipeline (#3909)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 138, "deletions": 3}}, {"node": {"oid": "e86214e5c00ffbb95b85478c111c8eb21de94fe9", "messageHeadline": "Fix the tensorflow performance test (#3847)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "7c774e967ad4df3b7fedcc1db3e036969ffe33b6", "messageHeadline": "support quantization of optimized model with ir<4 (#3853)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 15, "deletions": 11}}, {"node": {"oid": "25257a661d5543658b170bac93576800bf831216", "messageHeadline": "Added onnxruntime aarch64 wheel to pypi publishing pipeline (#3903)", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 64, "deletions": 2}}, {"node": {"oid": "1c1685ad2b916154250827f7eee030c54a5306ee", "messageHeadline": "Fix error handling in LearningModelSession.cpp (#3920)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "385073e1cdfaec932a229478a51b10dbb6f29c17", "messageHeadline": "Fix DmlCopyTensor test (#3923)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 11, "deletions": 20}}, {"node": {"oid": "eab61e87ce13cb5611a2f7a05779bfb20662a031", "messageHeadline": "Fix quantization tool bugs when model nodes have no name. (#3854)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "9b5daa2039c78d18a5ddb0021d47b6ff0ee605a8", "messageHeadline": "patch torch onnx opset 10 (#3910)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 212, "deletions": 4}}, {"node": {"oid": "7b858d60b0e3cf268908b7a972ff06e6e675dc24", "messageHeadline": "Various changes for automated downlevel test pipeline (#3901)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 281, "deletions": 291}}, {"node": {"oid": "3065219cc12ed83e090b58f8f45dd0e4cde46625", "messageHeadline": "Changes related to the release binaries requiring Visual C++ 2019 run\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 19, "deletions": 12}}, {"node": {"oid": "bccbdd03f10ad3d99261d6fa758667fea87ab798", "messageHeadline": "User/xianz/enable batch tests (#3914)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 26, "deletions": 1}}, {"node": {"oid": "18dc0ec39f867985b15ebe89b8e5f0ced90c4be0", "messageHeadline": "Rework jar by creating os-arch folders (#3849)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 115, "deletions": 39}}, {"node": {"oid": "c00945ae81fa93809ba3b337ed4738c087059457", "messageHeadline": "Build ORT by default for Mac OS X versions 10.12+  (#3626)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 24, "deletions": 7}}, {"node": {"oid": "99415f09fe86c4e078147c1130088facd93360b5", "messageHeadline": "Fix bug where linear_output_ is not cleared when linear_before_reset \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 47, "deletions": 35}}, {"node": {"oid": "475ea382f930bebaba9bb5be3c61b6d1d122de81", "messageHeadline": "Fix ACL EP convolution-activation fusion optimization (#3896)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "f170f31e1dbade5ee71d519a006051a6fae7612d", "messageHeadline": "Extend workaround with input name matching in DML fused graph kernel \u2026", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "0f82b42fede1afb663825615fd9804e926d6ad98", "messageHeadline": "Ensure pt model is set to cpu in ort_trainer (#3867)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 24, "deletions": 2}}, {"node": {"oid": "29496170d79b602c189b1c23665925860cebf518", "messageHeadline": "Add Benchmark Script for Bert Models (#3829)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 647, "deletions": 14}}, {"node": {"oid": "2fed37c8eb93d56ab00ad7c0af7921cf3abf934c", "messageHeadline": "Fix bug in handling of an initializer that provides a graph output. (\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 169, "deletions": 62}}, {"node": {"oid": "6f729b100f306d89769ef0d14d8eeb178828109e", "messageHeadline": "use LOAD_WITH_ALTERED_SEARCH_PATH for LoadLibraryExA (#3908)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 7}}, {"node": {"oid": "70abb120b3429b32477c719fe48b4264337813bc", "messageHeadline": "Remove ORTModel from frontend API (#3825)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 18, "deletions": 240}}, {"node": {"oid": "b12d35b6799bfbb432774d6f40b247c08bf479a7", "messageHeadline": "MLAS: tune softmax kernels for partial vectors (#3906)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 104, "deletions": 143}}, {"node": {"oid": "c46a9e8d65e2d9e5fdfbd777e37eb9ab73b25973", "messageHeadline": "Add numerical stability to SoftmaxGrad test inputs. (#3857)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 7, "deletions": 13}}, {"node": {"oid": "af7d45343589920fe6e3c9679d32489981f8b81a", "messageHeadline": "Merge DML Execution Provider updates (#3885)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 115, "deletions": 17}}, {"node": {"oid": "0cdc1459c64018ff61791e868876f0b7d38b9307", "messageHeadline": "Add a few more things to the helper python scripts. (#3842)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 278, "deletions": 22}}, {"node": {"oid": "28f693a8a13c3c1a3b563e7e8e75b633bae1f27a", "messageHeadline": "Update TensorRT dynamic shape profile when input shape changed during\u2026", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 11, "deletions": 10}}, {"node": {"oid": "6d2d927809a42e00ffbf4429cc8bd7bb2baa0d62", "messageHeadline": "add test to api tests to ensure correct hresult is returned for corru\u2026", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 17, "deletions": 1}}, {"node": {"oid": "c7da194313e35e78b2257d68d8cafa1c7b2635da", "messageHeadline": "remove winrt (#3899)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "ce3678ffafc5f057012e10d3eba3cabca2c69668", "messageHeadline": "Added aarch64 build pipeline (#3841)", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 88, "deletions": 4}}, {"node": {"oid": "d7e39569da75d19465b0358662d5f7ce9b0e7878", "messageHeadline": "Use MlasComputeSoftmax in traditional ML ops (#3892)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 35, "deletions": 9}}, {"node": {"oid": "06985a992226ba57036f503e709839e0b6504d0d", "messageHeadline": "Stop proceeding with constant folding if a CPU kernel is not found fo\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 56, "deletions": 44}}, {"node": {"oid": "7fd2c8f9e886298545f107dd8f274abce06ce63d", "messageHeadline": "Add signed GPU nuget package to publish ort-nightly nuget feed (#3834)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 40, "deletions": 0}}, {"node": {"oid": "5e1244eb4dcc5cea56e743a0071422c27c56848a", "messageHeadline": "Update ONNX submodule to ONNX 1.7 release branch. (#3888)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "96030fdcbccef153b8edaffd84051e33952cb2ad", "messageHeadline": "dashboard integration - output training perf metrics as json (#3809)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 459, "deletions": 24}}, {"node": {"oid": "8291065e209e725330edebdf8618d9c6c624b724", "messageHeadline": "Re-enable Dropout(12) ONNX inferencing tests. (#3833)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 3, "deletions": 6}}, {"node": {"oid": "eb33d5eda9acf3b66a205bf95267b1d730709f3d", "messageHeadline": "Do not register Dropout(12) as training ONLY kernel. (#3859)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 430, "deletions": 323}}, {"node": {"oid": "408f62dd5701ef3e06e734781b9257c6dfed5d97", "messageHeadline": "Load provider shared libraries relative to core runtime executable (#\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 35, "deletions": 5}}, {"node": {"oid": "3c248415693f75ae9690d0d4fe7504dd425e73ea", "messageHeadline": "Fold Shape Node During Constant Folding (#3748)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 314, "deletions": 91}}, {"node": {"oid": "890c945d4cf4d4ee69c177b02ce9227feb17d6ed", "messageHeadline": "winml_is_inbox is not accessible in function scope, move it out. (#3886)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 25, "deletions": 16}}, {"node": {"oid": "672e40bac21df501730917e879eef0355c4edc2f", "messageHeadline": "fix dnnl ep shared lib python packaging. (#3875)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "93e39b81a7edb1bd0fd63d608fd1b261e5c219ca", "messageHeadline": "fix build break on Linux. (#3874)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "cf6a1c1715643fb65fea4ecc44d4523b61f4ba6c", "messageHeadline": "Fix Windows Inbox build failing on 1) building raw api tests and 2) r\u2026", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 30, "deletions": 7}}, {"node": {"oid": "22a711457f369cef32b3d00881e9ed7da44a826e", "messageHeadline": "Fix C# log APIs. Also fixes github issue #3409. (#3840)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 91, "deletions": 26}}, {"node": {"oid": "a9b43b05cb51816db4aa671162088e87f6bafcc8", "messageHeadline": "Support more types in MurmurhashOp (#3827)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 46, "deletions": 7}}, {"node": {"oid": "19e3dc47d78582b2918249eddeede2bd1dfe7454", "messageHeadline": "LSTM performance tweak + cleanup (#3868)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 24, "deletions": 56}}, {"node": {"oid": "fe5a20f2d112fbaec79856f62bfee3a555e55bdd", "messageHeadline": "Fix initial value of graph optimization (#3856)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "eb7486d632c3fa58c845a7f6f91ad657fa7bbb70", "messageHeadline": "WinML Adapter DMLEP tests (#3752)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 390, "deletions": 0}}, {"node": {"oid": "e981496cf743a6148b1acae2bcc520d8798c7519", "messageHeadline": "add build inbox flag (#3855)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 15, "deletions": 7}}, {"node": {"oid": "424a00bf042b96b864b936324df69c4f2695dee4", "messageHeadline": "Fix enabling gradient as output for easy mode. (#3866)", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "4ea10c9202d024ae9b0d635ba4f8026f3569b539", "messageHeadline": "bump up ORT version and extend time limit for windows cpu packaging p\u2026", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "cb554fbc2da5ec24a108761c4c363f8d9781a38d", "messageHeadline": "MLAS: Add MlasComputeSoftmax/MlasComputeExp (#3846)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 3415, "deletions": 258}}, {"node": {"oid": "914aaaa1c84c4f576b39e1e80f0d27b23dec93d2", "messageHeadline": "Fuse Attention For One Input bert-base-dynamic Model (#3850)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 252, "deletions": 15}}, {"node": {"oid": "43b87de7d161ea71daab652d5ba154bb90e19b21", "messageHeadline": "Support wide char paths in CreateModel (#3835)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 14, "deletions": 3}}, {"node": {"oid": "3bafc096bb42ddbb9dc252d052ba35f44d642804", "messageHeadline": "Fast exit workaround for dnnl test. (#3863)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "0aeb383273ebc261d732a65d02351636f35c6f1c", "messageHeadline": "Support Pipeline in Training Runner (#3770)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 1123, "deletions": 169}}, {"node": {"oid": "c222ed6327fb984014ebca9d73a40246924d65ca", "messageHeadline": "Optimize Where CUDA kernel for UniLMV2. (#3799)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 207, "deletions": 150}}, {"node": {"oid": "65bfece19d98416d76b84bb42a534ff8f4b5f46c", "messageHeadline": "[Node.js API] optimize prebuild (#3844)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 125, "deletions": 753}}, {"node": {"oid": "9cca219b1aa2dd649d731298ba1249a184c9133f", "messageHeadline": "Add FAQ page (#3324)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 154, "deletions": 34}}, {"node": {"oid": "0e59668c1b8974afa9ec8e09e2da7b47dcbbb86b", "messageHeadline": "add support for symbolic broadcast for Add/Sub/Mul (#3743)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 108, "deletions": 21}}, {"node": {"oid": "687edd702cc45ffd1c969733e4488cc02d03b775", "messageHeadline": "Add RelWithDebInfo target to the C# projects so that it correctly fin\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 28, "deletions": 25}}, {"node": {"oid": "d38b79c6e59cabb029fa581e2c317573fb3f345d", "messageHeadline": "[Java] Adding missing methods on Session, SessionOptions and RunOptio\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 942, "deletions": 72}}, {"node": {"oid": "d5ec353e58114e18fa654c59262b5a3e63abd4c1", "messageHeadline": "Ryanunderhill/mkldnn dll (#3314)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1969, "deletions": 347}}, {"node": {"oid": "9b02b3df6f8f79951fecb1670a1da9a27fdf8bf1", "messageHeadline": "Update ONNX submodule to ONNX 1.7 release candidate 3. (#3838)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "b45ce925428efce1008b40adb9603cfb57500306", "messageHeadline": "Fix 3828 (#3837)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 31, "deletions": 31}}, {"node": {"oid": "f0c9fbc051ca1656161f06059ccc5fc763eecb88", "messageHeadline": "MaxpoolWithMask (#3831)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 205, "deletions": 97}}, {"node": {"oid": "edaf8a542c21b7f38cfc871fff9540ceb0e56951", "messageHeadline": "Initial PR for RKNPU execution provider (#3609)", "author": {"name": "airockchip", "email": "57983059+airockchip@users.noreply.github.com", "user": {"login": "airockchip"}}, "additions": 4034, "deletions": 2}}, {"node": {"oid": "584facf830644e2de3b9782a4bc48b26f716cfed", "messageHeadline": "Fix build error when DEBUG_NODE_INPUTS_OUTPUTS is on (#3826)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "6e341c002f6760baa5b08c6970d9cd994730a835", "messageHeadline": "Add Epsilon Attribute to Skip and Embed Layer Normalization (#3768)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 124, "deletions": 32}}, {"node": {"oid": "e30d2e38b92177cb251f9bcf2386185abb4004d9", "messageHeadline": "Add guidelines for writing a good PR. (#3830)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 16, "deletions": 0}}, {"node": {"oid": "ef4d73e88772bd99d1d35d05dc917e0a39906fe2", "messageHeadline": "Update ONNX submodule to ONNX 1.7 release candidate 2. (#3818)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "5db30a470e557c847c4ba1be9b5df2b74ca83c54", "messageHeadline": "[Java] Tidying up the sample MNIST code (#3824)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 116, "deletions": 21}}, {"node": {"oid": "f7ff5a7aa1813bddca716f07bfe8c0e6492cb8f0", "messageHeadline": "Fix state_dict and save_as_onnx for training (#3774)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 86, "deletions": 26}}, {"node": {"oid": "5dfc91db51c1c9fc63c046c8ff3dc682d9b7797d", "messageHeadline": "Node.js binding for ONNX Runtime (#3613)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 9486, "deletions": 0}}, {"node": {"oid": "cffa1b7bf22f95d07d59fa4bc87d6ddc2d9ebaa5", "messageHeadline": "Fix (#3812)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bd783644118e01cd00172c0fde3e4d9c386b1482", "messageHeadline": "Parallel all the activations ops (#3722)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1968, "deletions": 562}}, {"node": {"oid": "c11fbf68e438b35c5ea041365f7f8e6f968cea6f", "messageHeadline": "Publish gpu package to nuget feed (#3816)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 30, "deletions": 1}}, {"node": {"oid": "b386b41703f0d5f9467e23c4b905d85dc1ff7404", "messageHeadline": "Fix bug in GRU when linear_before_reset is true and no bias input is \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 83, "deletions": 40}}, {"node": {"oid": "a24c71af4091fb3f61088e455fdf54bd3c780c2a", "messageHeadline": "Update Dropout(12) forward kernel with training_mode input. (#3805)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 195, "deletions": 205}}, {"node": {"oid": "111469728f453e5eaf3e541a070661f193a22938", "messageHeadline": "Make Java build and run tests on Windows the box (#3811)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 45, "deletions": 7}}, {"node": {"oid": "6f95cdfa68d5d950f342f4f8ee83930cfbff8d67", "messageHeadline": "Use new cost based threadpool abstractions in CPU gradient operators.\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 27, "deletions": 30}}, {"node": {"oid": "156368b67f46fc6ab12b59ea39fcc93e51730ae7", "messageHeadline": "Quantize attention with Cuda (#3693)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 753, "deletions": 122}}, {"node": {"oid": "49f0610447fa3f111f263768fba121a12cce5d54", "messageHeadline": "Add options --disable_layer_norm,  --disable_gelu and --enable_gelu_a\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 52, "deletions": 3}}, {"node": {"oid": "2f8a2364c351d7a75dce01617d016095bc0a1b71", "messageHeadline": "Fix loss function builder (#3801)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "785b45124d6d231b2b0cb56813d38eec0fe726db", "messageHeadline": "Add CPU kernel for Einsum op (#3575)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2200, "deletions": 116}}, {"node": {"oid": "c8269e4b89813e9165d30e9507739ef826f2e612", "messageHeadline": "move backend test filters into data file (#3798)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 197, "deletions": 158}}, {"node": {"oid": "2684d47fc59993fb146ac44cdb93e3c23c2cff07", "messageHeadline": "Disable data downloading in linux-nocontribops-ci-pipeline (#3803)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 16, "deletions": 164}}, {"node": {"oid": "37b60251caebaef7fba768c5a6680dc2fa648701", "messageHeadline": "test packaging (#3756)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ee8900e21aa3cb41aa6482588180172c7ee730f6", "messageHeadline": "Update centos-ci-pipeline.yml (#3800)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "d5b2cd74939f79fa0e75736ace91f4962b66f3f4", "messageHeadline": "Add performance best practices to DML EP doc (#2859)", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "42cf971ca2a6700e7ca5a815282ded11da9aa71d", "messageHeadline": "Add a couple of utility scripts to tools/python (#3621)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 274, "deletions": 0}}, {"node": {"oid": "440f361363c0803af5d14e18e387d81803aaf832", "messageHeadline": "Remove orttraining-linux-gpu-inference-only-ci-pipeline.yml. (#3788)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 18}}, {"node": {"oid": "43a828f0a265882e616fa031879328eabc93bc9e", "messageHeadline": "Add tests for WinRT Projection Raw ABI consumption (#3718)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2216, "deletions": 3}}, {"node": {"oid": "3fab8ebfe9166fa03aece11eb98fb6a80b6d294d", "messageHeadline": "(MaximKalininMS) Fix Reshape Fusion and Crash in Reshape  (#3777)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 204, "deletions": 3}}, {"node": {"oid": "15eca74d152e68055b0881ffd601b134a276dc7d", "messageHeadline": "Make ThreadPool::PartitionWork a bit more user friendly. Update a few\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 44, "deletions": 47}}, {"node": {"oid": "2b8d9ef0fde81ea362d8b15577b4d839e804b228", "messageHeadline": "Refactor scatter/gather ops to use the new cost based threadpool abst\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 129, "deletions": 114}}, {"node": {"oid": "4f9f6aedea5709d579f2644b4f40953f0bc1865c", "messageHeadline": "CUDA/CPU test for NegativeLogLikelihoodLoss(12) function based loss o\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 110, "deletions": 0}}, {"node": {"oid": "b1c4d6ff4e7ebb2fa77595c3cbc718c1bfbb9d26", "messageHeadline": "bump dml version (#3792)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4b8fad214a18861a167ada995a8bc05626626e97", "messageHeadline": "Initial checkin (#3791)", "author": {"name": "David Brownell", "email": "38224104+davidbrownellWork@users.noreply.github.com", "user": {"login": "davidbrownellWork"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "11b819054b09ce2f404135f08f577f7359f82680", "messageHeadline": "Fix tree ensemble threading bug (#3778)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 131, "deletions": 71}}, {"node": {"oid": "2fc3984e70c83ad8c326afb6f75d5ed4dcd843e6", "messageHeadline": "Add test that C is unidirectionally broadcast-able before fusing the \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 60, "deletions": 17}}, {"node": {"oid": "e8e95110d33005b647d7389d35ab35dadb775c7a", "messageHeadline": "add pipeline to distributed context config (#3789)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 28, "deletions": 12}}, {"node": {"oid": "517bff9675b35306b74184802fe322285a3203b1", "messageHeadline": "Function expansion support and Update ONNX to 1.7 release candidate 1\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 186, "deletions": 69}}, {"node": {"oid": "047975e4040ef4d7f0d2d5e5b8fec66141f0a8f6", "messageHeadline": "Address flaky test ReduceApiTest.Sum. (#3716)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 88, "deletions": 83}}, {"node": {"oid": "edd5855fb7b7ed9e66c48d6fb2816ea7b4b8e068", "messageHeadline": "Remove eigen device from thread pool", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 33, "deletions": 59}}, {"node": {"oid": "dcb1a21552f8f3afdfff85bbc1c831e2a5684c88", "messageHeadline": "fix python package linux gpu failure (#3786)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "99ec93ea426ba2f286adb16212a6527987c46dc6", "messageHeadline": "Apply onnx-tensorrt bug fixes (#3785)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "e42e0d4787762dcb37e813009139d5167981e1ee", "messageHeadline": "Update documentation + Update mlas threading lib to use the new TrySi\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 15, "deletions": 15}}, {"node": {"oid": "29234458af083fe8bea743b6c899478717a53515", "messageHeadline": "disable cublasHgemm for training (#3769)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "98b97be635c0f2412724c8b62561d7caa537f7d1", "messageHeadline": "collect the last few iteration latency for throuput calculation (#3766)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "ad63e2593db278f34fc0e1ee6bbae926c62d8caa", "messageHeadline": "avoid using LocalFree on FormatMessageA buffer (#3772)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "f68a326bd94b711a68fbd98370f01e0174e04245", "messageHeadline": "Implement Pow(12) for cpu and cuda (#3727)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 507, "deletions": 107}}, {"node": {"oid": "027a364922d309a53548425188556ae70ac6b602", "messageHeadline": "Remove usage of openmp in reverse seq impl. (#3754)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 0, "deletions": 16}}, {"node": {"oid": "62c730a8df81c14657145022651e78ace167064a", "messageHeadline": "Revert softmax kernel implementation (#3753)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 346, "deletions": 204}}, {"node": {"oid": "86eaa71ec69fe8961dd58a364052d61831dae5ed", "messageHeadline": "sync threads before calling next cub function (#3758)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 12, "deletions": 1}}, {"node": {"oid": "af3988198cf173ad47b540ad26a045db7e38cdd1", "messageHeadline": "Liqun/e2e transformer test (#3540)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 728, "deletions": 610}}, {"node": {"oid": "177c1357f43ba3f933efb824521cc56c6414b014", "messageHeadline": "Use cublasHgemm \"back\" for fp16 computation with Volta GPU (#3765)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 31, "deletions": 21}}, {"node": {"oid": "3421ec1110e1338451ac649fce7ef5aca88ea141", "messageHeadline": "Add Threadpool::TrySimpleParallelFor  (#3759)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 35, "deletions": 12}}, {"node": {"oid": "b9a5ed1fe2b2fc58aab706a1967a6d659de70879", "messageHeadline": "Add SoftmaxCrossEntropyLoss to mixed-precision-transformer. (#3760)", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "9f72752397701028b7b46f269a1fa33e298d09f7", "messageHeadline": "Fix 'Install ONNX' CI failure (#3761)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 14, "deletions": 13}}, {"node": {"oid": "0531acccc510c3edd75c4f4a08d1f85db34475be", "messageHeadline": "Refine GatherND CPU/CUDA Kernels & Add UTs (#3688)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 236, "deletions": 143}}, {"node": {"oid": "58f53966d3e16b54c356aa323aac8a261a5826fb", "messageHeadline": "Add Distributed Checkpointing support (#3639)", "author": {"name": "ashbhandare", "email": "aibhanda@microsoft.com", "user": {"login": "ashbhandare"}}, "additions": 299, "deletions": 12}}, {"node": {"oid": "7296e06dd5f3689e5af8638847bf6c57cf2c1a26", "messageHeadline": "Properly creating arguments to pass to setup.py (#3744)", "author": {"name": "David Brownell", "email": "38224104+davidbrownellWork@users.noreply.github.com", "user": {"login": "davidbrownellWork"}}, "additions": 33, "deletions": 25}}, {"node": {"oid": "ea0e2d1ddebccaf25ba3948092e54bca94697420", "messageHeadline": "fix warning treated as error due to ignoring return status (#3739)", "author": {"name": "suffiank", "email": "suffiankh@gmail.com", "user": {"login": "suffiank"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "e529464a12ec975c9d63126bcde5215f73e436d8", "messageHeadline": "Limit the number of models run on OpenVINO (#3742)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "7ff06056bde9684ad416e0f66f31b3778d5207ff", "messageHeadline": "Fix the test coverage pipeline (#3710)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "0638565fe0d6149d1512ec3e63512e6ea13743a3", "messageHeadline": "Fix evaluation issues (#3538)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 243, "deletions": 31}}, {"node": {"oid": "939589c265d7046ac240ad04c18598200d37ef94", "messageHeadline": "Fix flaky test and avoid divide by zero in SoftmaxCrossEntropyLoss-CP\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 27, "deletions": 11}}, {"node": {"oid": "bad90d7a532cfe8e478b850c8a02a627da7e7376", "messageHeadline": "Fix a perf regression by providing a better estimate for the cost in \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "12d7c2f6e4eaab4f32a2ddba538641be4f45fb78", "messageHeadline": "iOS cross build on MacOS (#3699)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 126, "deletions": 40}}, {"node": {"oid": "29c12c0f07037ed38c684d28fc36ca24d4143508", "messageHeadline": "Handle dim with value of zero in ConvTranspose (#3728)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 41, "deletions": 7}}, {"node": {"oid": "9a4d1c772094ba2518b39d77207a2f852a9fedcb", "messageHeadline": "Merge pull request #3708 from microsoft/jeffbloo/MergeDmlDev", "author": {"name": "Jeff Bloomfield", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 3964, "deletions": 747}}, {"node": {"oid": "f1a948fd62da4a66b37a5e6d1717519042c15d32", "messageHeadline": "Enable telemetry on windows zip packages (#3738)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "78fde2c4cbeaaaf46fa97e1ffaea0fa01cbbcb33", "messageHeadline": "add downlevel test artifact to windowsai-nuget build (#3711)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "f7cf703d106ce59c01fa1b623c791c03673accd5", "messageHeadline": "[OpenVINO-EP] Optimize MCR Docker image size (#3732)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 39, "deletions": 57}}, {"node": {"oid": "1356215bd0483ce6fe329e8f5a4f98df0cc7d7de", "messageHeadline": "Fix build issues in the Python Packaging pipelines. (#3725)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 7}}, {"node": {"oid": "1bcfd499187af48f1f3b07e58f933b0f135c678e", "messageHeadline": "Merge pull request #3731 from microsoft/ettao/ort-2-master", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 926, "deletions": 950}}, {"node": {"oid": "6b3b4fe43e38c3c3f6f09dcad39988c804310b52", "messageHeadline": "remove warning message (#3730)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "1a11ba8a7eaefbb923a55f7f28cad147ab3fdc33", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into jeffbloo/MergeDmlDev", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 147, "deletions": 45}}, {"node": {"oid": "f487cc0b2835892dcd223b9eb9093f62e64ea294", "messageHeadline": "Fix Reshape Fusion with graph inputs (#3729)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 62, "deletions": 11}}, {"node": {"oid": "75c24a5facc535e1b28443523d3131089a760450", "messageHeadline": "Revert \"Merge from ort_training to master (#3719)\" (#3726)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 950, "deletions": 926}}, {"node": {"oid": "b990ba0059291addeb220e87a88ef9d2d7621baa", "messageHeadline": "Merge from ort_training to master (#3719)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 926, "deletions": 950}}, {"node": {"oid": "4f887b465a607fe559df882ac2a1673d17204f79", "messageHeadline": "Uncomment celu test. (#3717)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "7627e6bcc26893a96344ad90fabb11cd76530348", "messageHeadline": "Improve node and node argument name generation (#3649)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 79, "deletions": 26}}, {"node": {"oid": "407492472f66cbfe4a6cea61f985dcb0c57abade", "messageHeadline": "Fix build warnings and address PR comment", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "d03c552992bbee7ad3605fb32705605528ca0c83", "messageHeadline": "fix path of cpu_featurizers_kernels.cc and cpu_featurizers_kernels.h", "author": {"name": "Weixing Zhang", "email": "wezhan@microsoft.com", "user": {"login": "weixingzhang"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "635bc9cd0492504b86883327b62a7b12d22181ba", "messageHeadline": "Fix graph transformers to support opset 12 ops (#3715)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "0516e7d22e344a94bee021c791ddf80995b53ece", "messageHeadline": "Merge branch 'ort_public_ort_training' into ettao/ort-2-master", "author": {"name": "Ethan Tao", "email": "ettao@microsoft.com", "user": null}, "additions": 926, "deletions": 950}}, {"node": {"oid": "d9016408173171b0c439df93b0d05ee3e6a46596", "messageHeadline": "Call optimised version of depthwise ConvLayer (#3664)", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c23b484275794aa4826b5495202746bba3cc7a3d", "messageHeadline": "add missing deps in Dockerfile.openvino", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1621a1fef1523420c238f590baa22c4fcedbb3f8", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into jeffbloo/MergeDmlDev", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 941, "deletions": 140}}, {"node": {"oid": "e51c6c0b3b3a1893067688def31b1f552900c934", "messageHeadline": "Fix build warning in DmlOperatorResize.cpp and ReadbackHeap.cpp", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "805ffc01e5d5e3dd2edecd96dd8c41ee5a5ba13e", "messageHeadline": "Temp remove  --enable_wcos --use_winml from CI build (#3707)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "735caecfe12a4bd624759c078dce82e14451a87a", "messageHeadline": "Copy disabled ONNX backend tests from WindowsAI", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "a917023f94be664fb82db2adcfacce6f35e32f74", "messageHeadline": "Support for country-specific holidays in the DateTimeTransformer (#3701)", "author": {"name": "David Brownell", "email": "38224104+davidbrownellWork@users.noreply.github.com", "user": {"login": "davidbrownellWork"}}, "additions": 188, "deletions": 3}}, {"node": {"oid": "02e8d10f3a86c6ed54cf9678a989fdbe10192445", "messageHeadline": "Fix AdapterSessionTest", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "bf1caba2b2160d95f9b0c1cdb987537e91c014b6", "messageHeadline": "Port MLAS to Power architecture (#3703)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 750, "deletions": 134}}, {"node": {"oid": "f1c19f8495d0896ee5bdd66a079c49e34a1f0fed", "messageHeadline": "merge master", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 61316, "deletions": 5456}}, {"node": {"oid": "99a0bdf27178d4f518743f4ddd813338407c2c4e", "messageHeadline": "Upgrade nuget version in dml.cmake", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8cc161aec603443bd2de627e37cd3db75df7510d", "messageHeadline": "Remove problematic change for dxcore.lib", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 13}}, {"node": {"oid": "c49cc0c937e9125f9d48b31896a936f8ff427aff", "messageHeadline": "Increase DML nuget version to 0.0.2", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e22d97ba564c6789bf9ffeb5abb81ba0873c5cf4", "messageHeadline": "Merge pull request #3643 from microsoft/ort_training_for_merge_to_master", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 54394, "deletions": 1406}}, {"node": {"oid": "a475f2824df2fde34e9a204dac2f9e409d73c7d1", "messageHeadline": "Create the Nuget WindowsAI Pipeline (#3684)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 651, "deletions": 190}}, {"node": {"oid": "1c484ce33fe251cd817a1b14e8880f01fbd0cf66", "messageHeadline": "fix test (#3700)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "72b38f0a8bbb43c5d49905445e40d68def2c8366", "messageHeadline": "Support GPU Event Operators (#3653)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 317, "deletions": 87}}, {"node": {"oid": "8b5d6fbaf5aa490d7ad4234d80abac07b5771c14", "messageHeadline": "Remove internal work item links. (#3698)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "5637cec199cfc42241f5c4732a68e00654b1a97d", "messageHeadline": "Merge pull request #3697 from microsoft/ettao/merge-from-master", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 3831, "deletions": 2479}}, {"node": {"oid": "d06763ac1ca8c50dc78f74fcd756931c1e4c39f2", "messageHeadline": "Set gradient as output only for easy mode (#3694)", "author": {"name": "ashbhandare", "email": "ash.bhandare@gmail.com", "user": {"login": "ashbhandare"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "1b7bf481fcc09bc58930b1d4d9975ca894871ce3", "messageHeadline": "Ensure that setup.py works for all cmake generators and working dirs \u2026", "author": {"name": "David Brownell", "email": "38224104+davidbrownellWork@users.noreply.github.com", "user": {"login": "davidbrownellWork"}}, "additions": 25, "deletions": 3}}, {"node": {"oid": "e9f1e7e7972f1b2b659a7bbff4ef3f8bbbc9c72f", "messageHeadline": "resolve conflicts", "author": {"name": "Ethan Tao", "email": "ettao@microsoft.com", "user": null}, "additions": 3831, "deletions": 2479}}, {"node": {"oid": "4aa033b99e021d608f396bfd82ed098bf07c6dcd", "messageHeadline": "Addressing review comments (#3690)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 43, "deletions": 67}}, {"node": {"oid": "7347c7313961c00ae32decc926c3a3a0da42f657", "messageHeadline": "Revert \"resolving conflicts from master (#3691)\" (#3696)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 75}}, {"node": {"oid": "f48b9e2ea70ce8637cb145da2014cf66abe85048", "messageHeadline": "Add adapter session tests (#3522)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 415, "deletions": 26}}, {"node": {"oid": "c38a60a450d24a14a3b45301de2b9b6e8ca2096a", "messageHeadline": "resolving conflicts from master (#3691)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 75, "deletions": 7}}, {"node": {"oid": "63e6c257e4af0e8939f167484af944356e820a35", "messageHeadline": "Disable GeluApproximation transformer by default (#3644)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 32, "deletions": 2}}, {"node": {"oid": "ad8eb921d32f1e0e168afb535fde68284eda8e98", "messageHeadline": "Changes on RollingWindow Transformer (#3679)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 347, "deletions": 131}}, {"node": {"oid": "66343e2fcfef1be19d2af26429b6fb990e505f75", "messageHeadline": "use map with case insensitive hash and equals functions for learningm\u2026", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 33, "deletions": 2}}, {"node": {"oid": "5c7f6164310ffcf0a6021195ecae1e4a1654ecd9", "messageHeadline": "FeaturizersLibrary update and add variadic Input/Output to TimeSeries\u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 156, "deletions": 13}}, {"node": {"oid": "b4d4ea2e5f3322af47b3a2d952f3e2dbe326e3e8", "messageHeadline": "GatherND-12 Implementation (#3645)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 595, "deletions": 808}}, {"node": {"oid": "6d4f2f5bf92f86a8bbe52b6ecfed4bd1eb5e4ed1", "messageHeadline": "OpenVINO EP v2.0 (#3585)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 2842, "deletions": 2299}}, {"node": {"oid": "2f8a17dcde225373aa9886bd65e3e726147b2a35", "messageHeadline": "thrustallocator is not needed since cub is used directly for gather n\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 0, "deletions": 31}}, {"node": {"oid": "c929963d741d782d819e412006d45dec95ece7bc", "messageHeadline": "type cast for ratio is not necessary for dropout (#3682)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "f4a04c04e116ec976265b83a07dc95f467393caa", "messageHeadline": "move cpu/cuda related files to coresponding cpu/cuda folder (#3668)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 9, "deletions": 19}}, {"node": {"oid": "470f6e34d0f60d6d6dcc3443e04c6016fb3cbae6", "messageHeadline": "remove microsoft.ai.machinelearning.dll binpace (#3678)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 0, "deletions": 12}}, {"node": {"oid": "d97cb7338cf37e8d78cbe9f29c1f5c4020aed6cf", "messageHeadline": "Revert a change in attention", "author": {"name": "Changming Sun", "email": "me@sunchangming.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3863bd6f748b821bd4745eab1c11c9ce6ffb6ea2", "messageHeadline": "Revert \"Try not to modify base name (#3638)\"", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 47}}, {"node": {"oid": "2ab78c5da19ee64501c31f7b7e74e2fac44f22f8", "messageHeadline": "Update TensorRT parser (#3650)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "5a790a4b420cf14b8e55c17d4e8ca0f03f5536a8", "messageHeadline": "Merge remote-tracking branch 'origin/master' into ort_training_for_me\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 563, "deletions": 65}}, {"node": {"oid": "939d03666045f6e597b3f68e38a41d79400c5740", "messageHeadline": "Add omp impl for tryparallelfor and modify gelu to use fastgelu impl.\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 72, "deletions": 31}}, {"node": {"oid": "dd86e3be10b58f6240620dbc8146b0619f84ddbd", "messageHeadline": "Merged PR 4596882: Fix assert in ReadbackFromGpu", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "6ca44e216a5326bcc78d1d7e2768292bc74e2440", "messageHeadline": "Merge pull request #3675 from microsoft/edgchen1/merge_from_ort_training", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1599, "deletions": 209}}, {"node": {"oid": "2659f205cc597aeb997c87fb0b843b5b649ba73d", "messageHeadline": "Complex multiplication and conjugate contrib ops  (#3384)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 421, "deletions": 0}}, {"node": {"oid": "4416d4187430a6f5ae5b91cbace3dfd6b6d856e7", "messageHeadline": "Merge remote-tracking branch 'origin/ort_training' into edgchen1/merg\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1599, "deletions": 209}}, {"node": {"oid": "bae1dd7f0409968caf2c8f33038185878c2b3687", "messageHeadline": "add test for LearningModel creation from missing model path (#3661)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 15, "deletions": 1}}, {"node": {"oid": "b4e82913d1e26db993ebeafaf4f53c70fc2c894a", "messageHeadline": "Merge pull request #3670 from microsoft/edgchen1/merge_from_master", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 560, "deletions": 188}}, {"node": {"oid": "2d2375aa230b28a8bb56c217f1f1f5221e9fb2be", "messageHeadline": "swap float16/float (#3663)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c0e817ff161a558591180ab8d81412f39fa940d2", "messageHeadline": "Fix a bug in skiplayernorm fusion pattern 2 (#3660)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 55, "deletions": 18}}, {"node": {"oid": "deac467683dc2a42e603c8868b8659fe28ad69f3", "messageHeadline": "Merge remote-tracking branch 'origin/master' into edgchen1/merge_from\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 560, "deletions": 188}}, {"node": {"oid": "3ce31933bb9a653eae02af9ce55d328228b9dd10", "messageHeadline": "Wheel file updates for FeaturizerLibrary data (#3640)", "author": {"name": "David Brownell", "email": "38224104+davidbrownellWork@users.noreply.github.com", "user": {"login": "davidbrownellWork"}}, "additions": 51, "deletions": 15}}, {"node": {"oid": "ae7da2346028bc66b9cdcfaabb769cc51ae74cb1", "messageHeadline": "disable broken test in DML (#3666)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "49a1c5e546bce66c240490aabe64cce73f48b8e9", "messageHeadline": "Change CentOS build to use agent pool because builds on hosted agents\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "336624806e7acd3d3defb1b581b13341735cfdf6", "messageHeadline": "Simplify and clean code (#3655)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 31, "deletions": 38}}, {"node": {"oid": "125f68f305f3a345117b2b95174b7e62eec65654", "messageHeadline": "fixed mnist bug (#3569)", "author": {"name": "XiaocenDong", "email": "63833153+XiaocenDong@users.noreply.github.com", "user": {"login": "XiaocenDong"}}, "additions": 28, "deletions": 9}}, {"node": {"oid": "5777fc18c352589d093847277e35a9d4ddb170bf", "messageHeadline": "Removes omp for ThreadPool in TreeEsemble* (#3596)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 79, "deletions": 75}}, {"node": {"oid": "f1ba9aaf341c23bddeac9ac35659ea39eae9b32a", "messageHeadline": "Add pipeline transformer for wait/record node (#3513)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 551, "deletions": 113}}, {"node": {"oid": "6136fd0789c053783e304d168457d9f118e1d157", "messageHeadline": "GatherElementsGrad Kernels (#3627)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 656, "deletions": 48}}, {"node": {"oid": "d9641f292d474aaf8f428cce4547c7915ce60f5c", "messageHeadline": "Try not to modify base name (#3638)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 47, "deletions": 1}}, {"node": {"oid": "ffe19ae49b9820c1093ac400888d279a5a925ff6", "messageHeadline": "Expand elimination and Expand gradient. (#3610)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 285, "deletions": 0}}, {"node": {"oid": "37f4f74308ba856c866e589c99dae15dbdd62f7c", "messageHeadline": "expose training session so the training app could register custom ker\u2026", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "02bae6bd063024199ff241eacacc44d5db77490e", "messageHeadline": "Not use OpenMP for android build (#3636)", "author": {"name": "gwang-msft", "email": "62914304+gwang-msft@users.noreply.github.com", "user": {"login": "guoyu-wang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2dd4f7e96b5fa08802c3112f70d45821ac155b2d", "messageHeadline": "Add check for nullptr in PlannerImpl::FindReusableTensor(). (#3619)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "7c0b05eca0a4364bce7bca41c2accdfe5cabf750", "messageHeadline": "Merged PR 4593169: Handle empty tensors in DML EP batched tensor copy", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "0e12d05cd280e88ac0993fe830bc9f4e08efa2aa", "messageHeadline": "fixes for ort_trainer.py to resume from checkpoint (#3510)", "author": {"name": "suffiank", "email": "suffiankh@gmail.com", "user": {"login": "suffiank"}}, "additions": 145, "deletions": 67}}, {"node": {"oid": "00917917d6639827d99bfc34569d234be3f17f18", "messageHeadline": "Downgrade numpy requirement to 1.16.6 (#3635)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "e4fc83252d9d84285764590cdecbfa0915cafb9d", "messageHeadline": "Refactoring code related to WARP_SIZE. (#3623)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 128, "deletions": 163}}, {"node": {"oid": "3cf35955791299423d99dcceaa3909843485d73e", "messageHeadline": "Replaced spaces on tabs (#3555)", "author": {"name": "Mikhail Kuznetsov", "email": "61391551+kuznetsov-m@users.noreply.github.com", "user": {"login": "kuznetsov-m"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ab2b85649ef64699eeb4bb81c3bff34ade614bcf", "messageHeadline": "Merged PR 4587771: Merge Github master to DmlDev", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 65926, "deletions": 27545}}, {"node": {"oid": "26282359bf30d9f56acc14111bca6f73fc2c58df", "messageHeadline": "Merged PR 4591959: Fix ORT DML EP's Slice shape operator helper", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "7837c7efc358d2c492c47c2d0fe7421474b3abe4", "messageHeadline": "Add Features to ShortGrainDropper for ONNX export (#3628)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 166, "deletions": 32}}, {"node": {"oid": "bb9b0ba5b30230e8a48ca2e42ac6c46c61465cb4", "messageHeadline": "Merge pull request #3607 from microsoft/edgchen1/merge_from_master", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 11351, "deletions": 6001}}, {"node": {"oid": "70b554cc854003494b5c674aeb357df43ca0b509", "messageHeadline": "Add Features to ForecastingPivot Transformer for ONNX Export (#3608)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 241, "deletions": 56}}, {"node": {"oid": "ab70625b2981b0152cd87c6fc28533e01a2f0433", "messageHeadline": "Add Lamb shape inference (#3634)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 27, "deletions": 1}}, {"node": {"oid": "2c74766ad17214d1c2de0d5769a1809f4cd5b419", "messageHeadline": "Add new docs around how to bind to the onnxruntime.dll (#3539)", "author": {"name": "Paul McDaniel", "email": "paulm@microsoft.com", "user": {"login": "walrusmcd"}}, "additions": 13, "deletions": 0}}, {"node": {"oid": "8df5076d96fa566588b053d263c78274f6d52e1d", "messageHeadline": "Merge remote-tracking branch 'origin/master' into edgchen1/merge_from\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1128, "deletions": 1009}}, {"node": {"oid": "8d09cefafc4a122b9155731cb1f6188d885dd8a6", "messageHeadline": "Merge remote-tracking branch 'origin/ort_training' into edgchen1/merg\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 225, "deletions": 176}}, {"node": {"oid": "b518cb2a7a7fb511c7c1cbba72765d9d7f4b2a3f", "messageHeadline": "Clean up OPTIONAL name conflict workarounds in ort_training. (#3622)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 29}}, {"node": {"oid": "d3a2ac5c5cfdb78da5c9c8f770e940307f958af4", "messageHeadline": "Eliminate Useless Cast during Transformer. (#3606)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 136, "deletions": 0}}, {"node": {"oid": "d69bc313099a617d56803547cf561c383cfb4b91", "messageHeadline": "Refine BERT optimization script options (#3618)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 188, "deletions": 673}}, {"node": {"oid": "b4508dbdc64fea86bd438400d5f4f37219c8196a", "messageHeadline": "Improve TopK performance. (#3612)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 332, "deletions": 159}}, {"node": {"oid": "5492d02c4e0a88a70d3390f93fd517c2d5973889", "messageHeadline": "Remove Windows CUDA 9 build definition and helper scripts. (#3615)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 66}}, {"node": {"oid": "d66d5bb86acc4eea358ea08489aa0b3d583e6b00", "messageHeadline": "Update Optimizer Domain and Opset (#3602)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 88, "deletions": 81}}, {"node": {"oid": "47f1758fdc0219a591c512b0e7262f699a5f7fd0", "messageHeadline": "Add --skip_onnx_tests to orttraining Windows builds.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "297ab43b0c8080183e97ff7c5c41343b9a541b3f", "messageHeadline": "Add --enable_onnx_tests to Windows builds to allow set up of test dat\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "2e4b9b1d0e332abdddffe57df4d9f99ae6151bf3", "messageHeadline": "Disable CudaKernelTest.SoftmaxCrossEntropyLoss_LargeSizeTensor becaus\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "28a0c863b158033fc71c8bab11f20bede972fa61", "messageHeadline": "Revert \"Convert Gelu to use TryParallelFor (#3599)\"", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 29, "deletions": 14}}, {"node": {"oid": "d50c3e7a71617ca33e1d5cbc27cdc1f8bede6d4a", "messageHeadline": "Fix GraphTransformationTests tests.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 50, "deletions": 32}}, {"node": {"oid": "9636da39516cd07b2d5f53a2b04c9455186a9df2", "messageHeadline": "Threadpool related changes. (#3564)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 172, "deletions": 162}}, {"node": {"oid": "3dd3f84116148c1a5ae7a7dccf66b3826e5c47b2", "messageHeadline": "[Java] Adding model metadata support (#3573)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 417, "deletions": 10}}, {"node": {"oid": "c2a01b94312373def810497360ecfdaee877b4d4", "messageHeadline": "Disable erroneous compiler warning in space_depth_ops.cc", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "1c37d5e6ec730da6481436111bde142cf86b263c", "messageHeadline": "debug option for dumping tensorrt subgraphs. (#3604)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 19, "deletions": 5}}, {"node": {"oid": "87fad09c7b34a79311f6b120158e2c2cc0a6175d", "messageHeadline": "Fix merge issue.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "daa14b64e38b63b49f5b50129b4645dcef8cbcd0", "messageHeadline": "Merge remote-tracking branch 'origin/master' into edgchen1/merge_from\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 10266, "deletions": 5075}}, {"node": {"oid": "ead00f97f3fff8fa989dec1fba8ae97fadcbf78f", "messageHeadline": "Sync onnx_backend_test_series.py disabled tests (#3603)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 70, "deletions": 88}}, {"node": {"oid": "e233e6ba45f389abdb3fe69b46de2aa90e27ffbc", "messageHeadline": "Refactor - ScatterElements (#3559)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 62, "deletions": 55}}, {"node": {"oid": "2579a72a880cf72ced27b358db38b9f91d9132a9", "messageHeadline": "Convert Gelu to use TryParallelFor (#3599)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 29}}, {"node": {"oid": "971b98f9a5e5770c42f704f3b8a9ac770b7da738", "messageHeadline": "Fix ARM build error", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "911d1253239ada85af2f15efb24add00a2aa6c00", "messageHeadline": "Remove openmp from gpu build", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "850ab19e62df08bae46cdf6b5b461fe1015b3f79", "messageHeadline": "Fix Winml test build error", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "19cdd6f1e1d350d2b47c65f75706faa68c3d9903", "messageHeadline": "Fix chk build error", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "781e1c36bec18a1d48a639181ff0b7c682b8de8c", "messageHeadline": "Add front-end MNIST test (#3231)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 156, "deletions": 25}}, {"node": {"oid": "f180b71f27307d57d23c6693e30057a6e3772c06", "messageHeadline": "Support ONNX test version parsing from path on Windows in onnx_test_r\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 15, "deletions": 14}}, {"node": {"oid": "31b6629e993629c034541c37a760adde3c541079", "messageHeadline": "Fork WinML IDL Guids (#3591)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 55, "deletions": 54}}, {"node": {"oid": "381fee47ab7c6ef04ec2d152e4cd2650497b24a6", "messageHeadline": "Added support to build onnxruntime with ACL (#3586)", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 48, "deletions": 6}}, {"node": {"oid": "75426a30919b6dcf2a06883b37294d30f5521af6", "messageHeadline": "Fix build break", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "5d2874298e887c2958306e7e3fc252d413da3d07", "messageHeadline": "Merge remote-tracking branch 'upstream/user/jeffbloo/FreeDimOverrideB\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 90, "deletions": 31}}, {"node": {"oid": "88732cd09200b59b1c061f532409e31612488580", "messageHeadline": "upstream/jeffbloo/TrimOnSessionInitializationEnd", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 43, "deletions": 104}}, {"node": {"oid": "eceb18869a6e69a43767c0b37e12c71b73665150", "messageHeadline": "Merge remote-tracking branch 'origin/user/jeffbloo/BatchTensorCopy' i\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 141, "deletions": 13}}, {"node": {"oid": "acbfa42647691b2f1085802216d389b5a1b3fd6f", "messageHeadline": "Merge remote-tracking branch 'origin/DmlDev' into user/jeffbloo/Merge\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1090, "deletions": 431}}, {"node": {"oid": "7d523d258042ed43fc18264b0dac6a45b9462489", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into jeffbloo/TrimOnSe\u2026", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 10338, "deletions": 5024}}, {"node": {"oid": "414c4174a4165e06cf8f158063f818afdd640598", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/jeffbloo/Fre\u2026", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 8313, "deletions": 3986}}, {"node": {"oid": "8ee59531533fb7e5f003adbbc39864ee575b9a8c", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/jeffbloo/Mer\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 5545, "deletions": 1480}}, {"node": {"oid": "a4e312da43cb5be9d39fa6e011ef8763f34a251f", "messageHeadline": "Fix build error in D3DDeviceCache.cpp", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "422266c445201ae701b9674023d60711a008eee5", "messageHeadline": "Support conv transpos 1D in cuda provider. (#3300)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 24, "deletions": 9}}, {"node": {"oid": "7d5348f87e0ebaf748396d1c9230a2b0905fdccc", "messageHeadline": "Add ability to batch device copy for graph inputs and outputs. (#3580)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 97, "deletions": 15}}, {"node": {"oid": "ea62b3435a7c03eacf1ddd100bd18d1b17613e1e", "messageHeadline": "Clean up build.py code (#3466)", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 902, "deletions": 490}}, {"node": {"oid": "fcf0f6ee9f787d5713416904c539707dd59c4320", "messageHeadline": "Generalize reshape fusion (#3554)", "author": {"name": "Maxim Kalinin", "email": "makalini@microsoft.com", "user": {"login": "MaximKalininMS"}}, "additions": 213, "deletions": 92}}, {"node": {"oid": "14e387aa1a7d67e79502015f46dac3701796a615", "messageHeadline": "Fix WinML namespace build break (#3583)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 7, "deletions": 5}}, {"node": {"oid": "56b223bc60201dbea55950b7de0869ca1453d69b", "messageHeadline": "Implement OneHot CUDA Kernels (#3390)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 316, "deletions": 44}}, {"node": {"oid": "1599562016f736d892189ee207b4076dbe4cd201", "messageHeadline": "Fix BatchNorm CUDA kernel definition", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 28, "deletions": 35}}, {"node": {"oid": "c3658228086b8d11916e7a64d00cfd58d3fb3ddc", "messageHeadline": "Refactor some for the calibate.py. Add QLinearAdd and QLinearMul supp\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 89, "deletions": 15}}, {"node": {"oid": "db9566f70d2a1ca4ebebcf445a692dc835164293", "messageHeadline": "Implement Inverse(12) for CPU and CUDA (#3485)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 407, "deletions": 1}}, {"node": {"oid": "38a18023c77452ff9d19bee2b2d2ff74fe8359b5", "messageHeadline": "Fix some too popular warnings. (#3578)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 21, "deletions": 11}}, {"node": {"oid": "d68245853e439833f51bba6a7973680dbd6bad85", "messageHeadline": "Disable downloading test data on Linux (#3581)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 93}}, {"node": {"oid": "3e884b4b6b565575f39610f6881e7042991158a3", "messageHeadline": "Fix some typos. (#3582)", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 13, "deletions": 13}}, {"node": {"oid": "6fe688c73224c4ffa701dafcd41ad9ff3bd64db5", "messageHeadline": "Disabled failed maxpool test on GPU (#3549)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "52cfc98ec45fbf75db935d03bdc0d58588284831", "messageHeadline": "Merge pull request #3557 from microsoft/havenka/master-merge", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 5020, "deletions": 1575}}, {"node": {"oid": "811bd6787283652386b965ce3e50dba009ada34e", "messageHeadline": "Clean up docs. (#3579)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 84, "deletions": 130}}, {"node": {"oid": "ca1bbff5d4c2b995cf8fb16d6f57757653735352", "messageHeadline": "subgraph type override handling and unit test (#3560)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 184, "deletions": 35}}, {"node": {"oid": "e2288ff2b42545fb7f869be3bbd4639bc6ec5bde", "messageHeadline": "Merged PR 4574388: Implement Eyelike int ops", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dc576a8de8d24117bb3db7eabc96daecfee2cbf1", "messageHeadline": "Merged PR 4574316: Pad, OneHot, DepthToSpace, SpaceToDepth, TopK, Whe\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 24, "deletions": 17}}, {"node": {"oid": "c47490ab317a7db10a3eb7f5912768c712784dd5", "messageHeadline": "Bug fix", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c170d087a164cb23b37b10db359de6b8c6f6728b", "messageHeadline": "Merged PR 4573751: Add int32 and uint32 support to onnx Elementwise math", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "7f46f347db55a121e82f50609d4ce23b6301ed10", "messageHeadline": "Add GPT2 Attention Fusion in optimization script (#3488)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 965, "deletions": 149}}, {"node": {"oid": "5d3b21703923be12d149550826d7cb7054f5c1da", "messageHeadline": "Update Attention operator for GPT2 (#3474)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 185, "deletions": 42}}, {"node": {"oid": "2cb8cb816f2d281bc3bf1edcdd2c8a3e7cc6a7ef", "messageHeadline": "Disable or update flaky tests, improve test random seed accessibility\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 127, "deletions": 48}}, {"node": {"oid": "027b0cb3f3da2b250e6a9ef3de64bec6ccc486d6", "messageHeadline": "Update to match ORT signature", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 9, "deletions": 10}}, {"node": {"oid": "b4457ecb7ae84bbeb2d990be31f9aae5594b73a6", "messageHeadline": "Fix `gen_doc` build option and refresh documentation (#3545)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2280, "deletions": 494}}, {"node": {"oid": "5acd8dbe7da236e90db198a28bf09391ec6a2be7", "messageHeadline": "remove option --enable_lto (#3515)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "11baa4b56d4e0407a3d3f28e0296d2d08e64a624", "messageHeadline": "Merge user/ticastro/fix-wai-build", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 56, "deletions": 28}}, {"node": {"oid": "822883314e4e829c8f03db8d3e60b4dac82afa4b", "messageHeadline": "merge upstream/master", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 6475, "deletions": 3937}}, {"node": {"oid": "f822a548604e4e824f2269451aed15100444a0aa", "messageHeadline": "Make De/QuantizeLinear support half (#3531)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 360, "deletions": 128}}, {"node": {"oid": "c7b6fab29d9532a795f9d0a5f8ea68d8af1fe1e4", "messageHeadline": "Fix build break in mlas\\lib\\quantize.cpp: missing nearbyintf (#3572)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "eda4df62abd08f7fd5b954285df61bedba95923f", "messageHeadline": "Merged PR 4569464: ArgMin ArgMax Integer implementation", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "43c3a5edba79fd5de265955ae48b266e1c021a9f", "messageHeadline": "update onnxruntime version string for telemetry (#3526)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 2, "deletions": 8}}, {"node": {"oid": "de543c03088933fb4fc497e83ebc45484988358d", "messageHeadline": "Add SafeInt include to WinML targets (#3558)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "209b41a67d8b49a59ecfbd02fd334a7d5b51aed9", "messageHeadline": "Update dependencies graph", "author": {"name": "Changming Sun", "email": "me@sunchangming.com", "user": {"login": "snnn"}}, "additions": 88, "deletions": 21}}, {"node": {"oid": "2717c178cc63ac65c169539b8a35955afa153599", "messageHeadline": "Fork the WinML APIs into the Microsoft namespace (#3503)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2271, "deletions": 2215}}, {"node": {"oid": "fcb27c4e8bf2a65cdab4516109827228c881e170", "messageHeadline": "hotfix for skiplayernorm (#3543)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "92269ae409dd7192700e9d43eb981e4664048d0e", "messageHeadline": "perf tuning docs update (#3520)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "951484ba533a67e13e912a27b5950b01b9e0466e", "messageHeadline": "Dualapipartitionattibute.h header is missing in nuget package (#3350)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "1a222b3f6ea212221e03a73dec896aaf6b82ad47", "messageHeadline": "Disable downloading test data on Windows (#3551)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 34, "deletions": 138}}, {"node": {"oid": "c925156fd21129bd858055850508650592f5549e", "messageHeadline": "Fix build error", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d40fd897dda9840c9b4ed09b2a5ad8c86de0f8dd", "messageHeadline": "Support batched copies from GPU to CPU in DML EP", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 142, "deletions": 13}}, {"node": {"oid": "6aa9d14d02b8356f0ca4cc0ec3657c06c4de6611", "messageHeadline": "Address PR coomment", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "34bba7eac19d0ca1228d67dd9f94966916c6d053", "messageHeadline": "Merged PR 4568892: ORT DML EP OneHot build fix", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c7cefe920011bd541ee030c478569944a4d1fc31", "messageHeadline": "Merged PR 4558978: QLinearConv and QLinearMatMul implementation", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 416, "deletions": 183}}, {"node": {"oid": "be3b9244acba45889265ef8fa201e24aa696f1c7", "messageHeadline": "Fix GraphTest.UnusedValueInfoSerializes.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "9fc2b6482b0e6e77b84e8d0ccb5f3bcd63df4a72", "messageHeadline": "Ort training README (#3404)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 85, "deletions": 0}}, {"node": {"oid": "cc5a56516786d09f6806954b5540a863a759ec18", "messageHeadline": "Merged PR 4560079: MaxPooling dilations, TopK, SpaceToDepth, Mvn, Res\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 648, "deletions": 230}}, {"node": {"oid": "6c1ccb659fc39eb4223ab7695f1884c13fe48640", "messageHeadline": "SoftmaxCrossEntropyLoss-12 forward and backward kernel implementation\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 1435, "deletions": 14}}, {"node": {"oid": "757aa0fbaf6bd9d19ca6e8483d0444ecefd73606", "messageHeadline": "merge master", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 1845, "deletions": 1017}}, {"node": {"oid": "0ec90f7019178691a0ea13086d0ee94ec401d12d", "messageHeadline": "Put safeint_interface include directory into onnxruntime_common inter\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 24, "deletions": 22}}, {"node": {"oid": "80e0c64e2e4b68edfc547b26a70e200851195b80", "messageHeadline": "merged with master", "author": {"name": "harshitha", "email": "havenka@OrtTrainingDev0.af05slrtruoetgaxwwjv5nsq5e.px.internal.cloudapp.net", "user": null}, "additions": 5016, "deletions": 1573}}, {"node": {"oid": "93b957a55af37bf6a86d2acfe1ffbe417411b899", "messageHeadline": "Acl improvements (#3463)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 166, "deletions": 104}}, {"node": {"oid": "c91527235abff439dbaa2158bedbe45b3daeda3c", "messageHeadline": "[Java] Add support for map and sequence information on output nodes (\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 175, "deletions": 35}}, {"node": {"oid": "7c89f38a346bcfc4494884d8278ba838495e9575", "messageHeadline": "Fix static analysis warnings found by VC++ (#3530)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1496, "deletions": 873}}, {"node": {"oid": "47099a170072e2ced086b06c8d45e32172773793", "messageHeadline": "Merge remote-tracking branch 'upstream/master' into user/jeffbloo/Fre\u2026", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 354, "deletions": 77}}, {"node": {"oid": "65baa5aae31ebe7c222a529a2055ecfd090900a7", "messageHeadline": "Fix spacing", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a3a8a53736a475319f095dff805ab6d398cbf103", "messageHeadline": "Enable free dimension override by name", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 87, "deletions": 31}}, {"node": {"oid": "2f16172e69becc58b01ecff4036ec2d551a55fa3", "messageHeadline": "Address PR comments and clean up. (#3536)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "d18f0fc47b29a824de0750dfc132c2ead1bc232f", "messageHeadline": "Fix error handling for OnSessionInitializationEnd", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 5, "deletions": 3}}, {"node": {"oid": "ec4f6c099b3d3e2ea91106809b8979753af3a3e9", "messageHeadline": "Resolve comments and make minor changes to Featurizer transformers (#\u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 64, "deletions": 37}}, {"node": {"oid": "abfb275ac096209f176e2dc4cdf2b869393587a5", "messageHeadline": "Support listing keys in custom metadata map via C/C++ API (#3477)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 113, "deletions": 25}}, {"node": {"oid": "72cd61baaeecd8e27b5990caf61d7c98ac3598d9", "messageHeadline": "Removed use of parameters in python wheel build scripts (#3524)", "author": {"name": "David Brownell", "email": "38224104+davidbrownellWork@users.noreply.github.com", "user": {"login": "davidbrownellWork"}}, "additions": 17, "deletions": 10}}, {"node": {"oid": "cf2fddf76028274fe876c22785356c6bf1c7f7d7", "messageHeadline": "fix nuget build (#3532)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 40, "deletions": 0}}, {"node": {"oid": "2536e80602851a62132a536c0aa9f21563776ad5", "messageHeadline": "Rename API to onnxruntime.set_seed(<seed>)", "author": {"name": "Jesse Benson", "email": "benson.jesse@gmail.com", "user": {"login": "jessebenson"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "644bc05830fd6b03cd7779fb9d2396aeaa82b5f3", "messageHeadline": "Add Python API to set random seed:  onnxruntime.seed(<seed>)", "author": {"name": "Jesse Benson", "email": "benson.jesse@gmail.com", "user": {"login": "jessebenson"}}, "additions": 273, "deletions": 181}}, {"node": {"oid": "e89dd923875f8ad9ec66ea43db19cd637e0ccd1f", "messageHeadline": "Flush and trim resources in DML EP in new OnSessionInitializationEnd \u2026", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 41, "deletions": 103}}, {"node": {"oid": "b63349c8d6374ff7986587cdc6f57e4473f72069", "messageHeadline": " Fix custom op test failure (#3525)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bc9a199b165ac0b3502551fd1d3d5ac8cffc5a62", "messageHeadline": "Renaming deviceNum to deviceId.", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "e9dc8954ac85b70d3829796f0ef49910ea398b55", "messageHeadline": "Adding support for ACL and DML to the Java API.", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 68, "deletions": 2}}, {"node": {"oid": "2c7c45076b9a4a22a863b51c7737a7544ae6711e", "messageHeadline": "MaxBatchSize E2E Test  (#3454)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 71, "deletions": 2}}, {"node": {"oid": "a2feb29b0d76cd9816d03f25142fd394e15b6e5d", "messageHeadline": "Fix build break (#3528)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 52, "deletions": 3}}, {"node": {"oid": "1950391570752c04c25b99761ee51a69beeeb13d", "messageHeadline": "Specific DML version latest and fix compile error", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 5, "deletions": 4}}, {"node": {"oid": "e303f458e408e95735d428340056d328e0678679", "messageHeadline": "Add int64 input type for ReduceProd (#3507)", "author": {"name": "Negin Raoof", "email": "neginmr@utexas.edu", "user": {"login": "neginraoof"}}, "additions": 25, "deletions": 1}}, {"node": {"oid": "4fa88a0a23775dcc8bd70e16d201c6efc5f5a034", "messageHeadline": "Remove cast to OpKernelContextInternal to get threadpool and directly\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 6}}, {"node": {"oid": "f564569a8020e3c941c31e3f8404cfeb78903e6a", "messageHeadline": "Adapter Model and Environment tests (#3469)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 715, "deletions": 194}}, {"node": {"oid": "560f4c5b16ab52b05f34617080f1e3bf55c682d0", "messageHeadline": "Make GPUTEST macro consistent among TAEF/googletest (#3518)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 10, "deletions": 8}}, {"node": {"oid": "621b3ac03ac9117ecddde87b3c48dd2897f799c9", "messageHeadline": "FFT contrib ops (#3381)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 484, "deletions": 4}}, {"node": {"oid": "06b63975c0e026347da294454aa39fa6d75f3db3", "messageHeadline": "Fix fp16 type mismatch when graph output is an fp32-only node (#3411)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 25, "deletions": 2}}, {"node": {"oid": "baa86f181fec204a3d3bd35145facc5773cae3c2", "messageHeadline": "Handle the case that initializers are in graph input (#3449)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 58, "deletions": 8}}, {"node": {"oid": "006c5be1b17d9e9209ae443a522f8f7f5e62186c", "messageHeadline": "Optionally produce a python wheel that includes featurizers (#3491)", "author": {"name": "David Brownell", "email": "38224104+davidbrownellWork@users.noreply.github.com", "user": {"login": "davidbrownellWork"}}, "additions": 120, "deletions": 28}}, {"node": {"oid": "040c28ff3972347be32bb34b2a82771ac85a7d6e", "messageHeadline": "Remove dead code from HandleNegativeAxis", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ba7225f98651f146ec0b251833dafcd61990b6d2", "messageHeadline": "Update Graph SetInputs and SetOutputs for training (#3446)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 476, "deletions": 317}}, {"node": {"oid": "06db89cf13b4c5da0b9166de8422dde5af4ffdcc", "messageHeadline": "Using logic for finding README.rst to find requirements.txt", "author": {"name": "Colin Jermain", "email": "cjermain@gmail.com", "user": {"login": "cjermain"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "43d9f9190ea6c06493b4a392fd2823f3d558bb90", "messageHeadline": "Removing unused six package", "author": {"name": "Colin Jermain", "email": "cjermain@gmail.com", "user": {"login": "cjermain"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "c2c3102aba365b1890c358ccce1d479bea2ec1d1", "messageHeadline": "Tying install_requires to requirements.txt", "author": {"name": "Colin Jermain", "email": "cjermain@gmail.com", "user": {"login": "cjermain"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "66a79d2c9f71cd68d830b20af6e6abd5f73fb5a4", "messageHeadline": "fix (#3512)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "efd9b924824922e9f281e1859fbfecf963e176c1", "messageHeadline": "Handle Scalars in TernaryOps and Where. (#3509)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 36, "deletions": 26}}, {"node": {"oid": "cbe30f3e1970a3f35dd87b58ec19bb170052c45e", "messageHeadline": "update FeaturizersLibrary (#3511)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "5aab2671f8346a1489ac8ebf4b364544c6707062", "messageHeadline": "Fix crash in DequantizeLinear with scalar tensor (#3508)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 93, "deletions": 103}}, {"node": {"oid": "438353abcd3156e2273f5892f8452c5a2946280a", "messageHeadline": "Fix TruncatedSVDFeaturizer's test failure and re-enable it's kernel t\u2026", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 20, "deletions": 21}}, {"node": {"oid": "67443c0e17af8f712ae74c141a748ccd955bd65f", "messageHeadline": "Fix winml test compilation errors", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "5d99f179b9aad05d6fc5f1fe7565b40788709b21", "messageHeadline": "Merge pull request #3486 from microsoft/sedymche/merge_master_ort_tra\u2026", "author": {"name": "M. Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 6528, "deletions": 1806}}, {"node": {"oid": "d1384ba27e9f56d597466bab943b901399a47518", "messageHeadline": "Merge branch 'user/jeffbloo/MergeGithubMasterToDmlDev' of https://mic\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "e3c9fb9feed0040a50540067f727bb2217a13ce7", "messageHeadline": "Fix build issues", "author": {"name": "Jeff", "email": "jeffbloo@outlook.com", "user": null}, "additions": 3, "deletions": 4}}, {"node": {"oid": "4e045d39d355badc974b0eb2473a0710df7b1041", "messageHeadline": "Merge branch 'master' into user/jeffbloo/MergeGithubMasterToDmlDev", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 374, "deletions": 450}}, {"node": {"oid": "54bbbb78ae3eefe69bd27f71dcc5affa977842c2", "messageHeadline": "Change mask_index input of Attention op to be optional (#3459)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 133, "deletions": 46}}, {"node": {"oid": "bf3df414248684ca21f133da311aa7332617331b", "messageHeadline": "Put back SubmoduleCheckoutMode parameter into mac-ci.yml.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 3, "deletions": 0}}, {"node": {"oid": "b670cdc11d6984e7973e5c005da3f7cfdf31c00b", "messageHeadline": "Remove usage of DeviceProp (which is removed in ort_training) from cu\u2026", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 0, "deletions": 3}}, {"node": {"oid": "7b2fc196fe23dbcdbe82e20b638454ed95c8967a", "messageHeadline": "Get cudnn_common.cc from master.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 47, "deletions": 15}}, {"node": {"oid": "571a6d52f5d916f40789b7e227e71c3b58c750a2", "messageHeadline": "Add to list of failing backend tests from master.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 18, "deletions": 0}}, {"node": {"oid": "7f6e407e090da9ba42881d73c59b747312a41663", "messageHeadline": "fix python packaging manylinux1 build break. (#3482)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cffdff67020cf00caad1a40c907cf3bb2806c4ca", "messageHeadline": "Publish unit test results from Linux and Mac builds (#3480)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "f5ba9c922d64e5a9f75f30a37ab3294dee5d078e", "messageHeadline": "fix internal loss scale (#3483)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 41, "deletions": 9}}, {"node": {"oid": "4b5f66ad17f4433425439e10810470e844bff532", "messageHeadline": "Put dropout_default, dropout_random, celu back in the list of broken \u2026", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 3, "deletions": 0}}, {"node": {"oid": "42235910438edda51eba99008c5027ce1cde1ba9", "messageHeadline": "Add automatic generation of tensors for Onnxruntime Perf Runner (#3448)", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 38, "deletions": 1}}, {"node": {"oid": "56e85484ba2f8079a070c4c6e1f50fe024b1f4e7", "messageHeadline": "Handle optional inputs and remove more empty shape nodes in TensorRT \u2026", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 22, "deletions": 15}}, {"node": {"oid": "20c7dd9f5c20879d64edb9e5d18851f0c775b35e", "messageHeadline": "Remove orttraining/docker directory. (#3476)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 270}}, {"node": {"oid": "03996c7c0846020dbffe8daad93d801ceb22b5e5", "messageHeadline": "Fixes for Where, ConcatGrad and ReduceSumGrad (#3415)", "author": {"name": "Vincent Wang", "email": "wangwchpku@outlook.com", "user": {"login": "er3x3"}}, "additions": 208, "deletions": 28}}, {"node": {"oid": "d09d4a6b0df1ac103af4ea66a10b28b73d769bdb", "messageHeadline": "Fix OS build (#3481)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "95ade8f47bc257a152bab53dc44d9aae0b861cdd", "messageHeadline": "Add check to prevent storing nullptr in value_info_ when proto has un\u2026", "author": {"name": "Pranav Prakash", "email": "pranav-prakash@users.noreply.github.com", "user": {"login": "pranav-prakash"}}, "additions": 28, "deletions": 1}}, {"node": {"oid": "2ccedb7b4d55e00dbd77490cd057ede1991a3449", "messageHeadline": "Improve error logging when a kernel cannot be found. (#3473)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 34, "deletions": 14}}, {"node": {"oid": "bb2f4279908029e5d20646f0e2a07ef2d80d4032", "messageHeadline": "Remove FastGelu from activations.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 0, "deletions": 77}}, {"node": {"oid": "739c9d4875c91a069f246d4546f84d3e7a715ebc", "messageHeadline": "Always call cudaSetDevice at the beginning of session::Run (#3475)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 2, "deletions": 0}}, {"node": {"oid": "507d2bb9b93a7ac51dfb32633a8eb86e75d80ea1", "messageHeadline": "Get onnxruntime/core/providers/cuda/math/matmul_integer.cc from ort_t\u2026", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 0, "deletions": 23}}, {"node": {"oid": "1b465ba8206254eb1e386821d3a752b8aa1f9428", "messageHeadline": "Get onnxruntime/core/providers/cuda/cu from ort_training.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 23, "deletions": 116}}, {"node": {"oid": "c5176087bf006fc1631a4b817eee307d9f15c49b", "messageHeadline": "Get onnxruntime/contrib_ops/cuda/bert/fast_gelu.cc from ort_training.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 33, "deletions": 8}}, {"node": {"oid": "6bbc80951d3d3f984cb1c481ccc47dd67a85c7a8", "messageHeadline": "Get onnxruntime/core/providers/cuda/tensor/slice.h from ort_training.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "0e4080f1d696e19003d8ac9a99b057fedc7857db", "messageHeadline": "Get cuda_common.h from master.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 32, "deletions": 2}}, {"node": {"oid": "84773c61c6c5cbc671885c6e81e68eb5d6640fb0", "messageHeadline": "Rename ONNX OPTIONAL to OPTIONAL_VALUE.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "a443b1b6b9d15ccdacaaa6fa1d12fcda92d0cca0", "messageHeadline": "Revert \"Use IMMA for int8 matmul to leverage Turing Tensor Core (#341\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 113, "deletions": 371}}, {"node": {"oid": "e7297e6c9da720af093664d4da00744e4866e3af", "messageHeadline": "create pipeline for ci frontend tests (#3422)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 85, "deletions": 44}}, {"node": {"oid": "eaa3f652dfc75d95b6f71133cebc04acceadcdf4", "messageHeadline": "Fix dynamicslice.cc after merge.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 0, "deletions": 14}}, {"node": {"oid": "8ea0e596eca1151f86913f3f65c29799bda682c0", "messageHeadline": "Fix onnxruntime_unittests.cmake after merge.", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 2, "deletions": 0}}, {"node": {"oid": "6ba7c99e504ad5f2a48df1c73f2fb41754930d57", "messageHeadline": "Merge branch 'master' into ort_training", "author": {"name": "Sergii Dymchenko", "email": "sedymche@microsoft.com", "user": null}, "additions": 6578, "deletions": 1759}}, {"node": {"oid": "b04e48333d5f95580b961a61b2e370349ec4d205", "messageHeadline": "Merge branch 'master' into DmlDev", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 54448, "deletions": 22754}}, {"node": {"oid": "40d80cde8f26204ba6d43ab9f8ba34ebcfc73ca8", "messageHeadline": "Rework CDist (#3393)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 227, "deletions": 129}}, {"node": {"oid": "a08f16471a73b85d81f626dc9cfef213f0c49f81", "messageHeadline": "Address comments around bfc arena  (#3460)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "a298556600010c0e09e6522664eb4c77903bee01", "messageHeadline": "raid rtol to unblock CI (#3457)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 10, "deletions": 8}}, {"node": {"oid": "31bb1182e630e255268df5160a594a9058258086", "messageHeadline": "Merged PR 4530161: ORT changes for conversion of DML graph to public API", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 24, "deletions": 25}}, {"node": {"oid": "f73008483ac809e3e0348f0dfc564776610f404d", "messageHeadline": "safeint for region bytes in bfc arena and code clean up  (#3447)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 17, "deletions": 14}}, {"node": {"oid": "718068f02047370d647251763da28b8a172cfc31", "messageHeadline": "update C# API to optimize inference latency (#3171)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 1519, "deletions": 569}}, {"node": {"oid": "cdac74b3c3b2d21a5685016c53a70e0a7466d002", "messageHeadline": "Use Eigen threadpool for ReduceSum and ReduceMean. (#3441)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 8, "deletions": 12}}, {"node": {"oid": "1ddfe1249bd7e26940ff8a5319cc3722dc53e78e", "messageHeadline": "frontend test to use random seed (#3209)", "author": {"name": "liqunfu", "email": "liqfu@microsoft.com", "user": {"login": "liqunfu"}}, "additions": 32, "deletions": 24}}, {"node": {"oid": "f8fa1dde558ce2129ac27ba3bc5bca97866e4a74", "messageHeadline": "Add a list of Featurizers kernels (#3435)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 1009, "deletions": 7}}, {"node": {"oid": "4d71958ccff308de8f26eba8a7f5c72a895c1ab3", "messageHeadline": "Use IMMA for int8 matmul to leverage Turing Tensor Core (#3413)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 371, "deletions": 113}}, {"node": {"oid": "de60a14c16bf384d4a30c7e16a62ed2c89da333f", "messageHeadline": "Fix output range for int8_t QuantizeLinear op (#3445)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 67, "deletions": 78}}, {"node": {"oid": "aabf47b107a18efae3f7e318358352eb205b7907", "messageHeadline": "Fix Split CUDA implementation for zero sized input (#2942)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 37, "deletions": 26}}, {"node": {"oid": "48e96ea65f286eb5bf4e05bcc9b5011a18bec274", "messageHeadline": "Reduce binary size of Slice implementation (#3238)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 221, "deletions": 290}}, {"node": {"oid": "ff51d752d1d3f9ec317b69551ebf50c7fbc577d2", "messageHeadline": "Merged PR 4527632: Fix race condition creating resource store", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 4, "deletions": 6}}, {"node": {"oid": "b35468289ae1c78006aee4d622fdff80efbe0538", "messageHeadline": "View Op - new unit tests and add support for tensor memcpy by offset/\u2026", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 203, "deletions": 9}}, {"node": {"oid": "1fd21c109e2492678a6f7169b7d33e29245f5160", "messageHeadline": "Fix race condition creating ConverterResourceStore (#3419)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 4, "deletions": 6}}, {"node": {"oid": "53b9d52fc68f3bf7a1e167713067ef2268e677af", "messageHeadline": "Rework TensorToTensorProto.  Do not put string data to raw_string. El\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 15, "deletions": 14}}, {"node": {"oid": "43d6c464fc91e9b5dbc99738c42a013d14a6c375", "messageHeadline": "Fix ACL EP pooling build breakage (#3429)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 223, "deletions": 153}}, {"node": {"oid": "4bdb5cc8e2e536dfae50a3159a5d71b07defecb3", "messageHeadline": "Add CPU implementation for FastGelu operator (#3398)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 308, "deletions": 178}}, {"node": {"oid": "9e65298d7a320af0fa1e8d78b4f174f5c6b00a79", "messageHeadline": "Re-enable tests (#3437)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 30}}, {"node": {"oid": "15e32b44fd9dc1bbeb8aeb13522b876d89b9eac4", "messageHeadline": "Merge pull request #3383", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 9343, "deletions": 4683}}, {"node": {"oid": "8ab09186b74593a600dd7e103c74deaf3d97ce35", "messageHeadline": "Bert Optimization Script Improvements (#3387)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 107, "deletions": 48}}, {"node": {"oid": "95707d22a59b6f01170457814d4e4d0cbe6aeb4e", "messageHeadline": "Disable gradient clipping for E2E test.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "c8f5e6e632251821300237812401dbe7fa2bcf1c", "messageHeadline": "Implement Min/Max/Clip(12) (#3410)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 656, "deletions": 208}}, {"node": {"oid": "7c69b1703b16c6ab9a404c6068a1cbec84c61c81", "messageHeadline": "Fixed a typo (no functional change) (#3433)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "4ebad8805bc92513f3340e196d56061445c7d1db", "messageHeadline": "change (#3431)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 6, "deletions": 16}}, {"node": {"oid": "0dcc6035b134d31faae6ba41e6d1ebe54b43003b", "messageHeadline": "Disable strong inline (#3399)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 7}}, {"node": {"oid": "a3ab2ba03667cc857761940241b246760495cee7", "messageHeadline": "Reapply commit 131c65d; Fix memory regression issue. (#3423)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 46, "deletions": 39}}, {"node": {"oid": "d361121d9824b9f661c0f0f72079c45bf4d69dab", "messageHeadline": "Do not inline ExternOp's scalar tensor inputs (#3426)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 31, "deletions": 1}}, {"node": {"oid": "517693a507c31c976df0997bd1fb94dcf536987f", "messageHeadline": "Fix race condition creating ConverterResourceStore (#3419)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 20, "deletions": 12}}, {"node": {"oid": "33006f48c04953363a71ff43dbaf4894e73b2fbd", "messageHeadline": "Update onnx submodule to 1.7.0 release candidate (#3405)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 697, "deletions": 150}}, {"node": {"oid": "d4d19a75ba39aaf76fb7bc732a06244f87e7a01b", "messageHeadline": "Use MlasConv for 1D convolutions (#3425)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 28, "deletions": 3}}, {"node": {"oid": "58353496140d27465153579d1572d6c4fa229967", "messageHeadline": "Add #pragma once to providers.h, so avoid 'struct' redefinition error\u2026", "author": {"name": "Jesse Benson", "email": "benson.jesse@gmail.com", "user": {"login": "jessebenson"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "82c1e1b3db44013d80bb852bffea9f36fc4055c3", "messageHeadline": "Enable loss scale input from Python frontend (#3327)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 235, "deletions": 203}}, {"node": {"oid": "f4376653605bedee222c79a1335dc3cbec379f83", "messageHeadline": "Revert \"Addressing PR comments (#3334)\" (#3412)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 34, "deletions": 35}}, {"node": {"oid": "14f4c3e25fc79e47e2d02d5e039c10d6a60f97c6", "messageHeadline": "Fix issue in construction of DummyArena. (#3416)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 29, "deletions": 1}}, {"node": {"oid": "85131e760c374807653d004f468072e3fadbebb3", "messageHeadline": "Enable upsample2x optimization for opset 11 Resize (#3388)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 169, "deletions": 65}}, {"node": {"oid": "d89e5d91a63863cd4cdc179b56e3bd4272091fcf", "messageHeadline": "Disable GradientCheckerTest tests for GPU/Debug build (#3407)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 2, "deletions": 5}}, {"node": {"oid": "675035b1a8c7b6e1a19d7f7809fad0088938f5c1", "messageHeadline": "Disable GradientCheckerTest tests for GPU/Debug build (#3407)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 2, "deletions": 5}}, {"node": {"oid": "3568f8d186df180a2b942a42254dffd761ab2ea6", "messageHeadline": "Allow a custom op with the same name to be registered for several pro\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 40, "deletions": 6}}, {"node": {"oid": "a5fea26cb4fbb741f2cb49f6ffe7670b98455a0a", "messageHeadline": "Disable model tests for Mac OS X builds", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 27}}, {"node": {"oid": "e2afe5e054e1c5f7e294dd3b7e51453be8e2d741", "messageHeadline": "Revert Session and InferenceSession implementation", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 39, "deletions": 44}}, {"node": {"oid": "aefa466334f43a89c0821f74aef7f354f799e356", "messageHeadline": "Allow zero in split op (#3389)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 9}}, {"node": {"oid": "1671072b6ba2e51ba2e7891b7db5b6ed6cd8682b", "messageHeadline": "[WIP] Port image tests from WAI (#3365)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 1511, "deletions": 0}}, {"node": {"oid": "0b1e3f1e1012480587d42ad474613bbd4142ee0e", "messageHeadline": "Revert _SliceKernel cuda implementation", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 7, "deletions": 4}}, {"node": {"oid": "28ff88ce5259c6833c68428c036ffac76bdc34b6", "messageHeadline": "Disable tests (temporary)", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 82, "deletions": 4}}, {"node": {"oid": "1c334ed0f15f9013bfc2f1c424a224b0bd55f746", "messageHeadline": "Add Ninja generator to build.py (#3331)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "edec8043d464576f949ede0d3d80b1fbe22e663e", "messageHeadline": "Fix python examples in documentation (#3379)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 20, "deletions": 4}}, {"node": {"oid": "2ce90cff4c4ea04fb50d28aba3c2f5dbd4beaca5", "messageHeadline": "PR comments (#3374)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 57, "deletions": 46}}, {"node": {"oid": "614eb438aee493d673668b866657fa5f476d8519", "messageHeadline": "Update Op's Domain and Version (#3356)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 286, "deletions": 271}}, {"node": {"oid": "accffded5deaa561c400d1f8945ccd8dab589a35", "messageHeadline": "Build options for enabling AVX/AVX2/AVX512 (#3373)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 35, "deletions": 8}}, {"node": {"oid": "77c7d09ced384e030d6229a973713646c78f8bb6", "messageHeadline": "ERROR_NOT_SUPPORTED doesn't trigger Failed Hresult. Need E_NOTIMPL (#\u2026", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "052c1fda440533ee1be6c7a3e85390f8659421be", "messageHeadline": "fix some warnings in concurrency tests (#3395)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "33d3239b67371443eff5d819aaac1b060a6c56d1", "messageHeadline": "Rework SVMClassifier to improve performance (#3363)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 387, "deletions": 230}}, {"node": {"oid": "6d769d47c401cf818193edb00ef56a530a85cc1b", "messageHeadline": "Fix InferenceSession API", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 2, "deletions": 36}}, {"node": {"oid": "efc8bd738f91cd37dc5ee252355c0b0910b21b99", "messageHeadline": "add pipeline graph split script (#3275)", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 403, "deletions": 0}}, {"node": {"oid": "83c3da3fc046abe1231dbcb03a6176261323853d", "messageHeadline": "Fix code-base after breaking API changes", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 34, "deletions": 45}}, {"node": {"oid": "a61400de01130123c1acef4bd4d646cf723cc80f", "messageHeadline": "Fix ARM cross compilation (related to #3378, #3298) (#3385)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 14, "deletions": 15}}, {"node": {"oid": "ae15c36687294b83390bc8aaf806f68fa3411c93", "messageHeadline": "Merged PR 4500055: Fix autopilot WinML::Engine::Test::BenchmarkProtob\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 13, "deletions": 1}}, {"node": {"oid": "55fd283d20a91189659953ed268c142f8dbff570", "messageHeadline": "Fix a bug in FunctionImpl::FunctionImpl (#3376)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "a4fe60c4d3fb082d875b5371e7846f42916ea3b8", "messageHeadline": "OpSet 12 ops (#3341)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 992, "deletions": 149}}, {"node": {"oid": "044c4661581c3a0d3edd6dfc6181bca357306304", "messageHeadline": "Updated tags for v1.2.0 release (#3386)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "ecbacd7d79122472d51d9e4a90c6972070ac3bab", "messageHeadline": "Add Benchmark of GPT2 CPU inference (#3351)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1155, "deletions": 13}}, {"node": {"oid": "e6e4339f0bd2b5ad3087c5dfa9de1fb66dc03857", "messageHeadline": "Merged PR 4408731: Perf: Command lists should be preemptively reset i\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 13, "deletions": 14}}, {"node": {"oid": "759818f2c1799a7607deb3ba0edd5795105f8a7e", "messageHeadline": "Merge remote-tracking branch 'origin/master' into thiagofc/ort_traini\u2026", "author": {"name": "Thiago Crepaldi", "email": "thiago.crepaldi@microsoft.com", "user": {"login": "thiagocrepaldi"}}, "additions": 9301, "deletions": 4673}}, {"node": {"oid": "ace741680d6c8bd542fc2ba2bcb3aded40d9d319", "messageHeadline": "Constant-12 support  (#3304)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 570, "deletions": 54}}, {"node": {"oid": "2332a93db06dd91697f7cd0fe52b8eff68166457", "messageHeadline": "Update onnx-tensorrt parser (#3369)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "ce9acf0c2112a5af1107fae1e597b528a1b4c65d", "messageHeadline": "iOS crosscompilation under linux (#3298)", "author": {"name": "Jan Scholz", "email": "radikalliberal@gmail.com", "user": {"login": "radikalliberal"}}, "additions": 78, "deletions": 14}}, {"node": {"oid": "fb2f97a002209ea06a785a56bcce0a9a3dad175b", "messageHeadline": "Address master merge PR comments (#3348)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 80, "deletions": 43}}, {"node": {"oid": "af618278f6f8e1fae42a6035782f206640cdaeed", "messageHeadline": "fix bugs in quantization and calibration tools (#3329)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 63, "deletions": 19}}, {"node": {"oid": "f2ca2b2981c9081dc372f8e52b79dfb83035afd1", "messageHeadline": "Avoid \"infinite\" loop in optimizer (#3321)", "author": {"name": "Maxim Kalinin", "email": "makalini@microsoft.com", "user": {"login": "MaximKalininMS"}}, "additions": 24, "deletions": 13}}, {"node": {"oid": "06fc9506fdf83082f1ef226324fab31806dce1ce", "messageHeadline": "Thread pool changes (#3153)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3281, "deletions": 1566}}, {"node": {"oid": "0494036006993737bd6055a860e02627acd3a9cb", "messageHeadline": "fix tensor location mismatch in allocation planner (#3249)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 378, "deletions": 8}}, {"node": {"oid": "2b10e625f9c13a9e6e3b04aeabcab81c1630fc26", "messageHeadline": "added public value varibale to NamedOnnxValue (#3347)", "author": {"name": "Cassie", "email": "46505951+cassieview@users.noreply.github.com", "user": {"login": "cassiebreviu"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "355f39ddee885237bf6fde9b587c7a5f80d22c53", "messageHeadline": "fix cuda build for cmake >= 3.17.0 (#3362)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 6, "deletions": 1}}, {"node": {"oid": "d8f0a0f223ee2cda068158e5390e55efa800e9c3", "messageHeadline": "Address PR comments (#3352)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 64, "deletions": 3}}, {"node": {"oid": "1bbc421884e060bb0219912b3df7390ac02d1911", "messageHeadline": "Don't cast to fp16 in LayernormGrad (#3328)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "33b5010e624d0c649e121a323eda29dfa4074e2a", "messageHeadline": "skip optional inputs for scan subgraphs (#3349)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "324166b9cd295cc417a02d492282c6f3aa442392", "messageHeadline": "Merged PR 4471419: DML EP kernels in ORT", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1242, "deletions": 223}}, {"node": {"oid": "5bc81b7ae9a7de4b9caf2ea756a5b0e732701c55", "messageHeadline": "Fix bad merge (caused Slice to fail).", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "5972ce5566fc270db590145ab5218a4b5c973053", "messageHeadline": "PR feedback", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "f1a062c292955e200101fe7331076e959f85c0f6", "messageHeadline": "PR feedback", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 90, "deletions": 22}}, {"node": {"oid": "ffb2a3359e54817e683eb91e128a4bf4ad782922", "messageHeadline": "Implement WhereGrad (#3343)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 60, "deletions": 8}}, {"node": {"oid": "351c3c30fbc1d8cb5365fa4ef3f87f527fa15cc0", "messageHeadline": "Merge branch 'DmlDev' into user/dwayner/DmlEpGatherScatterReverseRang\u2026", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 623, "deletions": 117}}, {"node": {"oid": "8ff351ecc42845704c87fc848d1a85d4b8e85e5e", "messageHeadline": "Merged PR 4482575: DML EP Slice in ORT", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1281, "deletions": 116}}, {"node": {"oid": "6c960a9417447a09227b8a71ae735f3ddbc747b5", "messageHeadline": "PR feedback.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "c3cea486d06756a8c45708189372bf1b4ee5b1f7", "messageHeadline": "Port ConcurrencyTests from TAEF (#3086)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 514, "deletions": 30}}, {"node": {"oid": "49e6043d07fa1e48b5c65817edb3577988b69de2", "messageHeadline": "support Huggingface's adamw (#3318)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 1556, "deletions": 1282}}, {"node": {"oid": "031647635b4672bf67a2df277434faf7fa016005", "messageHeadline": "Delete another litter file.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 0, "deletions": 345}}, {"node": {"oid": "5feb3c0f19ecf30ad492324cebc1a46a00c60e1f", "messageHeadline": "Delete litter backup files.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 0, "deletions": 4943}}, {"node": {"oid": "89df6abac7ca4e80e396bd7ef4ed7768c3af6b1e", "messageHeadline": "Fix slice.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 6438, "deletions": 15}}, {"node": {"oid": "ccb840ac99ff3cea6c9e9da03d10e3118a913021", "messageHeadline": "Fix slice.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 126, "deletions": 119}}, {"node": {"oid": "5278f73202757547482534eab685ecaf2084f18b", "messageHeadline": "Fixed two issues in symbolic_shape_infer script (#3332)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "810a10b23054cf97daaa661a6cd0a38f6b3a6952", "messageHeadline": "Enable Onnxruntime Telemetry by Default for 1.3 (#3338)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "131c65d23d040079d50a6dc3e85f72f0de401346", "messageHeadline": "Addressing PR comments (#3334)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 35, "deletions": 34}}, {"node": {"oid": "2e875f4e6794b224ff96a887248e42635d11ac84", "messageHeadline": "Delete outdated page (#3320)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 0, "deletions": 88}}, {"node": {"oid": "497e83eda5c49d1613fa7e4b169b656193d7dec0", "messageHeadline": "Minor update to the issue template. Add a line to attach model where \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "0a6ec0df5699a3a3e7b94375b4d09c346d3942f4", "messageHeadline": "Merge pull request #3285 from microsoft/xuzhu/merge_from_master", "author": {"name": "Xueyun Zhu", "email": "40807589+xzhu1900@users.noreply.github.com", "user": {"login": "xzhu1900"}}, "additions": 9592, "deletions": 3003}}, {"node": {"oid": "d143b41b8175ac227bbe24778ae51d8594bee14b", "messageHeadline": "Expose frozen_weights in PyTorch Frontend (#3317)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "0e81962e98bc1e7dd1bb5efcd4414bfa53e133a4", "messageHeadline": "correct the cmake version to 3.13 for Arm build (#3333)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "13dabd97b6da9b79c1ddb69ec864ec9139133b23", "messageHeadline": "Slice", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 20, "deletions": 1}}, {"node": {"oid": "66c7579c9357291e5c16b4e7f8a543127e2f2226", "messageHeadline": "address PR comments (#3312)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 16, "deletions": 20}}, {"node": {"oid": "5f6ec8ea6d23abfeeda0ad185291d92fa40a046f", "messageHeadline": "Fix a bug in Maxpool v8", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 8}}, {"node": {"oid": "dee4fc8b8a58a72cc50fd17e36401f1f4619adea", "messageHeadline": "Apply the same check for no_transpose from the Reduce* ops to ArgMin \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 32, "deletions": 12}}, {"node": {"oid": "51e95ea9464e3ee4bb02b7ba2339266af1c54f77", "messageHeadline": "Make ort errors appear in winml exceptions (#3316)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "a031af8cfcb499d371d635657305e15fe46fc326", "messageHeadline": "fix windows build", "author": {"name": "Xueyun Zhu", "email": "xzhu1900@gmail.com", "user": {"login": "xzhu1900"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "9eb792a5b393aeffd794419a8195f2f6698a3e94", "messageHeadline": "move env to .cc file", "author": {"name": "Xueyun Zhu", "email": "xzhu1900@gmail.com", "user": {"login": "xzhu1900"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "b38fc0d541c433ec0dc43ecbd53fc18d578d4393", "messageHeadline": "Add bias correction in Adam & Lamb for C++ frontend & python frontend\u2026", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 221, "deletions": 36}}, {"node": {"oid": "0f6ace2c3649255beab55760bee7f95fddf384c3", "messageHeadline": "Fix comments with stale todo's.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1, "deletions": 6}}, {"node": {"oid": "55d32085f615634f8e7230b076c0dd5fced9c553", "messageHeadline": "Update comments.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 29, "deletions": 43}}, {"node": {"oid": "cc095fefbb884822cf809753548d358c7af348e4", "messageHeadline": "Fix squeeze.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 61, "deletions": 24}}, {"node": {"oid": "e9877850a48c5b17b4b688450ff902fde4a667d0", "messageHeadline": "fix python error", "author": {"name": "Xueyun Zhu", "email": "xzhu1900@gmail.com", "user": {"login": "xzhu1900"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "4db01309cb5fcf48d7e15986e0c6ffe44fcac55a", "messageHeadline": "Use GEMM for SVMRegressor. (#3305)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 181, "deletions": 53}}, {"node": {"oid": "5366273110bc7219970ee856d965713cbab5c54f", "messageHeadline": "Fix GatherND and ScatterND.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 287, "deletions": 137}}, {"node": {"oid": "6474801cebfb5c660a88e3d8eebabdb54ac933dc", "messageHeadline": "Update ort_trainer.py with lazy onnx export (#3244)", "author": {"name": "Bowen Bao", "email": "semisqg@gmail.com", "user": {"login": "BowenBao"}}, "additions": 94, "deletions": 52}}, {"node": {"oid": "19edad132cc256e49b95e88d2fcacba0d47ba122", "messageHeadline": "Move AzureML Bert notebook from onnx tutorial (#3302)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 648, "deletions": 0}}, {"node": {"oid": "fef79898662907acbc534ffb306c4b259aa032c8", "messageHeadline": "Replacing CudaAsyncBuffer with TArray to improve perf (#3303)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 424, "deletions": 358}}, {"node": {"oid": "98c28060b0cb88af8c535632b7b5332519485930", "messageHeadline": "Aggregated Send/Recv (#3232)", "author": {"name": "Li-Wen Chang", "email": "30609447+liwchang@users.noreply.github.com", "user": {"login": "liwchang"}}, "additions": 340, "deletions": 172}}, {"node": {"oid": "271126a86eaf2d5ed5b5d24f3ee4961838130171", "messageHeadline": "Fix clip.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 72, "deletions": 19}}, {"node": {"oid": "1df264f4b371bd990084871c8801a40c746ba012", "messageHeadline": "Fix Range.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 360, "deletions": 138}}, {"node": {"oid": "ef7b98f9886639dd8ed750532c132a5d96b2aadd", "messageHeadline": "Support DisposableNamedOnnxValue inputs in c# Run() (#3175)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 163, "deletions": 13}}, {"node": {"oid": "fb5ab858d26f9ad7a7b983dc9dfa477f3dfbabcf", "messageHeadline": "Update BUILD instructions (#3282)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d15c74e713b74cb9a5d34112f33d34bab5c4301c", "messageHeadline": "Implement pipeline event generator (#3206)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 241, "deletions": 42}}, {"node": {"oid": "8f7bd51f7af19388ccfa5ecfdf7d2478bb0f54aa", "messageHeadline": "fix pybind issue introduced by merge", "author": {"name": "Xueyun Zhu", "email": "xzhu1900@gmail.com", "user": {"login": "xzhu1900"}}, "additions": 55, "deletions": 48}}, {"node": {"oid": "b72fe13941e0ebd9c5de0b58e4b41b59cf60e544", "messageHeadline": "Update WinML Projection to accept sequence of tensors (#3287)", "author": {"name": "Sheil Kumar", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 615, "deletions": 202}}, {"node": {"oid": "7f610caca07d26094413172f53dcbe24b978a740", "messageHeadline": "Make gradient clipping configurable. (#3243)", "author": {"name": "Tixxx", "email": "tix@microsoft.com", "user": null}, "additions": 28, "deletions": 7}}, {"node": {"oid": "843ee346a83e7c0180d444df74eb487efa12b17d", "messageHeadline": "Implement struct TArray and simplify code. (#3291)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 30, "deletions": 26}}, {"node": {"oid": "57468c651c5b9bfc9a0f37c548203891e088f71a", "messageHeadline": "QLinearMatMul speed up (#3283)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 76, "deletions": 47}}, {"node": {"oid": "9c3b6d2e4b3f52964829aa7017bf912ea11f5e42", "messageHeadline": "Fix warnings in nuphar", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9dbc50c4388df988ef052e1add10e143570a66de", "messageHeadline": "fix build break", "author": {"name": "Xueyun Zhu", "email": "xzhu1900@gmail.com", "user": {"login": "xzhu1900"}}, "additions": 63, "deletions": 67}}, {"node": {"oid": "403f99cd77e596b8d161c2f3d2859548bc0d3bb2", "messageHeadline": "Use yapf to format python (#3276)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2121, "deletions": 2045}}, {"node": {"oid": "84015d9491bfd421f93295842dfab4123ed8bf3b", "messageHeadline": "Fix post merge test. This doesn't get triggered as part of gated PR c\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "ccc3535e72bac1fe36bcfcc6f9ee2e6266014b71", "messageHeadline": "resolve conflict", "author": {"name": "Xueyun Zhu", "email": "xzhu1900@gmail.com", "user": {"login": "xzhu1900"}}, "additions": 9498, "deletions": 2916}}, {"node": {"oid": "b880c48c4c999043c887289b288072393ecf30bb", "messageHeadline": "Make reduction ops handle Scalar input (#3260)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 158, "deletions": 24}}, {"node": {"oid": "5c5233222326e5e2970b217c3acb89d9870b93c5", "messageHeadline": "unittests comments (#3278)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 18, "deletions": 7}}, {"node": {"oid": "c5149e89d9392f1443ad283a813c5bf1391145a3", "messageHeadline": "Wangye/shortgraindropper (#3273) (#3274)", "author": {"name": "Ye Wang", "email": "52801275+wangyems@users.noreply.github.com", "user": {"login": "wangyems"}}, "additions": 278, "deletions": 45}}, {"node": {"oid": "1d9be2baed03a39e27b9a26df8b889cc2b098ce3", "messageHeadline": "Add Notebook for Bert Model exported by Keras2onnx (#3271)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1016, "deletions": 1}}, {"node": {"oid": "a69d8599121b30cab7497368ba77f64c8a536847", "messageHeadline": "fix quantize_bias (#3270)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 27, "deletions": 17}}, {"node": {"oid": "d521efd9046b0e825050beec3bdd0fb8e6142350", "messageHeadline": "refactor frontend (#3235)", "author": {"name": "liqunfu", "email": "liqun_fu@hotmail.com", "user": {"login": "liqunfu"}}, "additions": 386, "deletions": 294}}, {"node": {"oid": "d9f628cb1de3d48f43496c283a406bf3c2451eb5", "messageHeadline": "Remove orttraining/tools/scripts/profile directory. (#3268)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 1267}}, {"node": {"oid": "6dc25a60f8b058a556964801d99d5508641dcf69", "messageHeadline": "Make the reduction ops more consistent in checking if no transpose is\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 61, "deletions": 15}}, {"node": {"oid": "ca7985fd9fb481d0d87c49ac2ee4c1a6f44c0fee", "messageHeadline": "Address PR comments (#3256)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "8f00147c14c64715ffd4b1512df5356ddeb75462", "messageHeadline": "Fix a few warnings", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 14}}, {"node": {"oid": "3bdb0b620aaf726a67049886fec44b623dbc85b2", "messageHeadline": "Fix WCOS/Win32 linking bugs (#3126)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 175, "deletions": 109}}, {"node": {"oid": "61e8a24340e7fed6af4385844ef97a6af6796aba", "messageHeadline": "Address PR comments (#3255)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 76, "deletions": 87}}, {"node": {"oid": "d82f72e65c2d83c9afc53a2f884e988cc48d8c99", "messageHeadline": "Add ort_training build status file. (#3257)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 45, "deletions": 0}}, {"node": {"oid": "435f014d71436f63527a939ece9b4c460e2a8bba", "messageHeadline": "Add support for sessions to share a global threadpool. (#3177)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 936, "deletions": 233}}, {"node": {"oid": "03d14bae2b8ee928bc9c84958fa392f077e3c600", "messageHeadline": "Register ONNX Training Ops (#3252)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 4, "deletions": 6}}, {"node": {"oid": "e03b8a1e2ff92da1d8d3a20fe7279d31437f409b", "messageHeadline": "Move path_lib from onnxruntime/core/framework to onnxruntime/core/pla\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 18, "deletions": 14}}, {"node": {"oid": "61621d405325f8dca6de228592fdc886ea6d3967", "messageHeadline": "Add extra fields to ORT telemetry (#3234)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 32, "deletions": 12}}, {"node": {"oid": "bd348ec6ca6bab53e5be7870722ff6866b240383", "messageHeadline": "Add unit test to cover TreeEnsembleClassifier applied to binary class\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 58, "deletions": 0}}, {"node": {"oid": "88c65f8add003f77e1749a5321095e0fcfbfd0e7", "messageHeadline": "Fixes GTest deprecation warnings", "author": {"name": "jaka.katrasnik", "email": "jaka.katrasnik@sensum.eu", "user": {"login": "katrasnikj"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "c5576d70a647031b570be1de32d4f7e7155edee7", "messageHeadline": "Fix build issues (#3214)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 205, "deletions": 173}}, {"node": {"oid": "0700d13ece5f65bb8ce158fb94e836129240cde8", "messageHeadline": "Add Bert Optimization Notebooks (#3204)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3327, "deletions": 96}}, {"node": {"oid": "f6211217d8102998599a56847bac7ef30b101ecf", "messageHeadline": "Fixes", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 175, "deletions": 334}}, {"node": {"oid": "8bc4e3195df7c4e28d2b663656072f08db1ab7f1", "messageHeadline": "Updates to roadmap (#3155)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 11, "deletions": 7}}, {"node": {"oid": "e63f817eb67944c656c97769f7b3cd46e3800dfa", "messageHeadline": "avoid IDXGIFactory 6 where possible to enable WinML GPU Path downleve\u2026", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 45, "deletions": 18}}, {"node": {"oid": "682dde2b3baf9b62909c2b5de147898996479926", "messageHeadline": "add dml_ep_lock (#3200)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 41, "deletions": 3}}, {"node": {"oid": "4b2c8e884e258af619ce13240820f8b299082212", "messageHeadline": "Udpate License Header (#3212)", "author": {"name": "Sherlock", "email": "baihan.huang@gmail.com", "user": {"login": "SherlockNoMad"}}, "additions": 113, "deletions": 13}}, {"node": {"oid": "6319357a999a81bac39b799b223c6f0ed853720a", "messageHeadline": "Reduce number of allocations in TreeEnsemble (#3217)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 74, "deletions": 87}}, {"node": {"oid": "0fceb33288ce35472d1cbab24fd7d95d5d3c9b07", "messageHeadline": "Fix onnxruntime server docker file build failure (#3219)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 242}}, {"node": {"oid": "3a7539e0716f590c5350b379670c82cd3726a632", "messageHeadline": "Update bert-base convergence values", "author": {"name": "Jesse Benson", "email": "benson.jesse@gmail.com", "user": {"login": "jessebenson"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "dc11b82956e498a99374e03abf5fe6c0e31874da", "messageHeadline": "Tweak the dropout calculation.", "author": {"name": "Jesse Benson", "email": "benson.jesse@gmail.com", "user": {"login": "jessebenson"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "88c20eaef17c6a5b0b4f3e4277edfe0f969c52de", "messageHeadline": "MLAS: rename AVX512BW->AVX512Core (#3216)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 79, "deletions": 76}}, {"node": {"oid": "551d28be9a162422f41487a76bc653b2e60b9b4a", "messageHeadline": "Update.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d489288e3cc1b97cffb9451e644a8b51e23cc860", "messageHeadline": "Add kernels, including stubs.", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 1393, "deletions": 67}}, {"node": {"oid": "2a6e5ce97818771a6aac62cac1b52e10ae1baf41", "messageHeadline": "Speedup and reduce binary size for TfIdfVectorizer (#3197)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 202, "deletions": 298}}, {"node": {"oid": "fe0b2b2abd494b7ff14c00c0f2c51e0ccf2a3094", "messageHeadline": "QLinearConv speed up (#3196)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 573, "deletions": 319}}, {"node": {"oid": "0a1257e46731d10edb25b59c852b97c07555ad5e", "messageHeadline": "Adjust the grouping logic in ThreadPool::TryBatchParallelFor (#3207)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 30, "deletions": 6}}, {"node": {"oid": "5bc0d8be5c328f56c017e3006dc7941cd672191d", "messageHeadline": "Fix TopK Cuda implementation (#3176)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 19, "deletions": 7}}, {"node": {"oid": "93569bf0f4f596d2b700d27e3560ee4a08c2dfac", "messageHeadline": "fix regex to populate dll version information correctly", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "c69194ec4c8c9674368113aa6044d0db708cd813", "messageHeadline": "fix the missing return in _get_quantize_input_nodes and format code w\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 611, "deletions": 291}}, {"node": {"oid": "d99554bea16c27d62f34ca6705efad3b74d0d3a8", "messageHeadline": "Improves implementation of tree ensemble regressor and classifier (4 \u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1264, "deletions": 707}}, {"node": {"oid": "e9d5ed270fadd9225bdaedba715f6c5dd3e1ebb5", "messageHeadline": "Normalizer performance improvements (#3201)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 126, "deletions": 132}}, {"node": {"oid": "890cb78b2022ad1fe702de8f1484f1ebdc1cb871", "messageHeadline": "Use Eigen::logistic instead of manually computing values. (#3186)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 22, "deletions": 7}}, {"node": {"oid": "b8575dda7be471af14607de0230767d38f79606e", "messageHeadline": "Avoid some heap allocations in the InferenceSession and Model classes\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 210, "deletions": 185}}, {"node": {"oid": "24793f5fc7de6d588bd97a904d993b81a6a24403", "messageHeadline": "Revert change from RelWithDebInfo to Release in OnnxRuntime.CSharp.sln.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "2cad08bd6061d9d6915081ce058ee275a935374c", "messageHeadline": "Merged PR 5688: Upgrade ONNX submodule to the latest from github ONNX\u2026", "author": {"name": "Zeeshan Siddiqui", "email": "mzs@microsoft.com", "user": {"login": "codemzs"}}, "additions": 1025, "deletions": 114}}, {"node": {"oid": "2f1e997e5b60d495888f5273e8eabb98bf1a1544", "messageHeadline": "Merged PR 5686: fix P100/fp16 issues", "author": {"name": "Ethan Tao", "email": "ettao@microsoft.com", "user": null}, "additions": 96, "deletions": 22}}, {"node": {"oid": "75025461e2981e3f461226a2a6c63f43f58fba96", "messageHeadline": "Initial implementation of graph cut and pipeline", "author": {"name": "Ke Deng", "email": "kedeng@microsoft.com", "user": null}, "additions": 756, "deletions": 147}}, {"node": {"oid": "a02638eb4660c70ed44d0439be6a6d31cb7530cd", "messageHeadline": "Adjust the threading logic in ThreadPool::ParallelFor (#3178)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "f49912c42a6467c36676b84490cca8b22851fdc9", "messageHeadline": "Performance improvement to Transpose when moving single axis. (#3173)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 40, "deletions": 42}}, {"node": {"oid": "fa4dd51e3b0ef33c354127b6f4f442d26d0424e6", "messageHeadline": "Add back orttraining-linux-gpu-inference-only-ci-pipeline.yml. (#3182)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 18, "deletions": 0}}, {"node": {"oid": "3af5a2a2cfeecd5fe3cb0c8445688abe4f229cc1", "messageHeadline": "Change Tensor::[Set]ByteOffset() to use ptrdiff_t.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "80dd62a240494fb7aa9b4ba700ccf1798547f99d", "messageHeadline": "Enable CI for training.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 71, "deletions": 258}}, {"node": {"oid": "e542cfd0e074cafab2b9e3c5dae9a5884546649d", "messageHeadline": "Introduce training changes.", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 48487, "deletions": 739}}, {"node": {"oid": "6791ed0217849fd129bc618d0015d1e2bdad2b8e", "messageHeadline": "Documentation updates for 1.2 for WinML (#3149)", "author": {"name": "Paul McDaniel", "email": "paul_mcdaniel@hotmail.com", "user": {"login": "walrusmcd"}}, "additions": 140, "deletions": 26}}, {"node": {"oid": "a912415bac038a9b49733129f8bee3c042988d84", "messageHeadline": "Support custom ops targeting the CUDA EP (#3165)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 56, "deletions": 21}}, {"node": {"oid": "3464801c3e2e3a4bee0cfdf313ab98c5b7a207d2", "messageHeadline": "Explicitly specify NugetPackage parameter while validating nuget in s\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "3de1fc096d2fd8c770c96f7ed4cdc6fa9aa55f9b", "messageHeadline": "Move zero point inputs of MatmulInteger to CPU memory (#3159)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 27, "deletions": 29}}, {"node": {"oid": "51a8c829086b9378d68ae7ca2b5097c51743da5f", "messageHeadline": "Update bert optimization script for SQuAD model exported by keras2onn\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 571, "deletions": 53}}, {"node": {"oid": "be0c0da9eff3805cc9b0543d99d077826b0d2391", "messageHeadline": "Merged PR 4408735: Fix control over metacommands being enabled or dis\u2026", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 42, "deletions": 11}}, {"node": {"oid": "876d0c543028cb9e43c2c319d7541428b716ab09", "messageHeadline": "Make quantization parameters as constant weigth instead of overrideab\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 0, "deletions": 14}}, {"node": {"oid": "3d928de7781b3e82790a1a6ac01661f0c9a49023", "messageHeadline": "Use GEMM for LinearRegressor and LinearClassifier operators to improv\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 483, "deletions": 219}}, {"node": {"oid": "f87b6913cd504df87b7ed71986369a833a4e9425", "messageHeadline": "Add package download step before pushing to feeds (#3162)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "6ed5d7c332920343b7147debbf77870faf738fcc", "messageHeadline": "Update post_binary_sizes_to_dashboard.py (#3161)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 7}}, {"node": {"oid": "a59243090a7c71732c1d7e681c2aeffe9e10c8ec", "messageHeadline": "Publish release symbols (#3152)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 24, "deletions": 9}}, {"node": {"oid": "781a6ebb06fd4311705628f97c363d181feee35e", "messageHeadline": "Updated Ruby supported versions", "author": {"name": "Andrew Kane", "email": "andrew@chartkick.com", "user": {"login": "ankane"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cfd18b583a4f0a1d382c6c20e730acc99da8cf6b", "messageHeadline": "Help output typo fix", "author": {"name": "pranavm-nvidia", "email": "49246958+pranavm-nvidia@users.noreply.github.com", "user": {"login": "pranavm-nvidia"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5be6665b8603f0409926f93f0336680fd7ac7ce6", "messageHeadline": "Update Gelu Fusion to support new graph pattern from PyTorch 1.4 (#3148)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 332, "deletions": 47}}, {"node": {"oid": "e2894c5ffbd993b5b1ff61ac751f244fa443e90d", "messageHeadline": "Fix package name overrides (#3150)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 10, "deletions": 6}}, {"node": {"oid": "1d2b8115e2e61bec87a114214f328b2985cd797c", "messageHeadline": "Support u8u8 in quantization tool (#3140)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "ade4fa108f2061734d7e9fd6f48f5e1991b46bc2", "messageHeadline": "Disable delayload for cuda dlls (#3147)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 16, "deletions": 6}}, {"node": {"oid": "2c446a7f2f4aa2d9e087908b4e326dc9aa92055f", "messageHeadline": "Add push to ORT-NIGHTLY. (#3146)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 17, "deletions": 1}}, {"node": {"oid": "fbb658e6034f5cfb6823df5e4996538ba919e2ec", "messageHeadline": "Implement QuantizeLinear and DequantizeLinear (#3098)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 261, "deletions": 0}}, {"node": {"oid": "83753bcbe3b9a2f9571378f2e6a244d086a2770f", "messageHeadline": "Suppress maybe uninitialized warning in gcc-9", "author": {"name": "take-cheeze", "email": "takechi101010@gmail.com", "user": {"login": "take-cheeze"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ef8768a53f2346d41e21d47b97a2986f893db064", "messageHeadline": "Override native package name. Preserve managed package name the same.\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 6, "deletions": 32}}, {"node": {"oid": "a2eeb126b9553827cce4a401b8a857df78d4a49e", "messageHeadline": "Optimised kernel_dot() in SVM op (#3135)", "author": {"name": "Prabhat", "email": "prabhat.roy@microsoft.com", "user": null}, "additions": 24, "deletions": 19}}, {"node": {"oid": "9d874c1225befe52eae1d0953df6cf80a44d82af", "messageHeadline": "Add bert performance and correctness test tools (#3108)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1036, "deletions": 9}}, {"node": {"oid": "84ad4eda8b336b733d36caeb1d5924531d465d09", "messageHeadline": "Implement MatmulInteger on GPU (#3070)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 584, "deletions": 0}}, {"node": {"oid": "12605f05d14f29a8741432d32db983a3c89c7c9e", "messageHeadline": "Fix CUDA PATH (#3131)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 7}}, {"node": {"oid": "6cdd2b4934ef42518f80d5b0938c08285100d22e", "messageHeadline": "Enable DML Nuget Package for x64 or x86 architectures (#3120)", "author": {"name": "smk2007", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 158, "deletions": 46}}, {"node": {"oid": "e45326b5dfec934d3deb9b3fea628d678dbde2c3", "messageHeadline": "Create NuGet packaging pipeline for ORT Featurizers (#3125)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 103, "deletions": 13}}, {"node": {"oid": "b538cb7e466f7a4fb4cec2b2084366a2818fcaaf", "messageHeadline": "NCHWc Upsample/Mul optimizations (#3116)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 452, "deletions": 76}}, {"node": {"oid": "4188b1111a8680dd56124c9ea7d55373a14987a2", "messageHeadline": "Add a summary for each ExecutionProviderAppend methods in SessionOpti\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 36, "deletions": 11}}, {"node": {"oid": "37f5fd8fb8981e38c836005a1845d422552017b2", "messageHeadline": "Add support for loading TensorProtos with external data from optimize\u2026", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1212, "deletions": 245}}, {"node": {"oid": "c6ed0774417aa1efb2b2bdd99f08f051a43cd682", "messageHeadline": "Add d2FH4- flag to cuda (#3105)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "86b755774f86fcdb18cdcea1af6e6a991f3af3b3", "messageHeadline": "Create a separate Nuget hosting just managed assemblies (#3020)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 476, "deletions": 210}}, {"node": {"oid": "5008fc5b00e9808b276466aeaaca29f6620b521d", "messageHeadline": "Featurizers: Import fix for Linux build adjust linkage (#3089)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 12, "deletions": 27}}, {"node": {"oid": "d72639ef775a23a910d6ea19cfde9367204330b2", "messageHeadline": "Fix CUDA 10.1 DLL names (#3102)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "37a905f557cd3b16c89c75e7b9e81002447b4431", "messageHeadline": "Make Java API available on Android (#3030)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 145, "deletions": 17}}, {"node": {"oid": "ca2ed17ba75389338fe63e85af2adcd5f49fb15c", "messageHeadline": "Bump up version number to 1.2 (#3097)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "f1ba531d9c40c4f511ad5f19909a7c1a69aac615", "messageHeadline": "Disable test_zfnet512 and test_bvlc_reference_caffenet for x86 in C# \u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "04063aa324f3e445c77e4d29a0f563da9397c417", "messageHeadline": "optimizer shouldn't depend on inference session", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "bf7afbef235daf94302334aace9d03b29f4a6cbe", "messageHeadline": "Changes in the props file to support .NET + AnyCPU configuration (#3091)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "5e0f7412cdb754f814e402fc0e24fc732dba6997", "messageHeadline": "Properly handle downlevel and WCOS scenarios (#3075)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 62, "deletions": 36}}, {"node": {"oid": "57384d50779453b4fe64504f241c4d9945f5a81f", "messageHeadline": "Add disclaimer to DataFrameTool header (#3085)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "d7500b26bd65e3b3a850d55994cec56ad76bdca0", "messageHeadline": "Remove Publish Build Symbols from pre-checkin  CI build (#3088)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 7}}, {"node": {"oid": "f4a5d172941d8e2034a3e93cf60cb5fbb9dfab0f", "messageHeadline": "Upgrade to CUDA10.2 for TensorRT (#3084)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 65, "deletions": 65}}, {"node": {"oid": "d7f2cdcc7ef8ef95a2d3d2af9d3c528256162471", "messageHeadline": "Fix target platform of managed OnnxRuntime dll and enable x86 .NET te\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 55, "deletions": 20}}, {"node": {"oid": "9ba05a4a1cc8b37ee70ccf1c3651e92313d60c02", "messageHeadline": "Merged PR 4349264: Updating 8 Bit Kernels", "author": {"name": "Nick Feeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 181, "deletions": 24}}, {"node": {"oid": "92d8d90d24c521ddd707b244f9f560337dd1d02c", "messageHeadline": "Comments", "author": {"name": "nickfeeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 21, "deletions": 19}}, {"node": {"oid": "f407de6da1df8c3597ab575d9b6544bc4cc4d655", "messageHeadline": "Updating 8 Bit Kernels", "author": {"name": "nickfeeney", "email": "nickfe@microsoft.com", "user": {"login": "nickfeeney"}}, "additions": 178, "deletions": 23}}, {"node": {"oid": "b23b7f0fea6c12d149d1825f49b78ababd61eabf", "messageHeadline": "[java] Adds the provider compile-time flags where the JNI code expect\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 34, "deletions": 8}}, {"node": {"oid": "a506911208e1b9e5da23de1f432b45acccc8c46d", "messageHeadline": "No need to create a copy of graph proto when checking to see if there\u2026", "author": {"name": "Ryan Lai", "email": "rylai@microsoft.com", "user": null}, "additions": 8, "deletions": 10}}, {"node": {"oid": "dae9a31719388177d9f66ca71f9f6c59951e4ccf", "messageHeadline": "Introduce new Featurizers packaging pipeline. (#3068)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 148, "deletions": 0}}, {"node": {"oid": "b8628404f3ae711d5d10c20c86d5e8e31132672e", "messageHeadline": "Replace hardcoded include path value with the advertised setting. (#3\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "fb7317173d58b7317ff2cac889d4f13d5a2a46af", "messageHeadline": "Doc updates for 1.2 release (#3069)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 26, "deletions": 24}}, {"node": {"oid": "0d30b42c7bc89ec07611cd763e170e68d218cb35", "messageHeadline": "Fix python gpu build (#3081)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f367fd921c81840f7039a308aeb9c9a8eef4d60e", "messageHeadline": "Use a custom allocator for temporary buffers in reduction_ops.cc (#2775)", "author": {"name": "kile0", "email": "kile@microsoft.com", "user": {"login": "kile0"}}, "additions": 129, "deletions": 32}}, {"node": {"oid": "61ae134469eeb5d31919376019098fd4a40c4821", "messageHeadline": "Fix binary size report (#3080)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 17}}, {"node": {"oid": "3e371575ea876dc05f7edeb10ad9222a181bc596", "messageHeadline": "Relax testing tolerance of NchwcOptimizerTests::BatchNormalization (#\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ae1f35fb9fce25162dd73b73d3061db313b6d9f9", "messageHeadline": "Ignore GCC no-deprecated-copy warnings (#3074)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a0c9f679fc27b295f6ff5f2cb8276665e95871f8", "messageHeadline": "Add logging to nchwc_optimizer_test.cc (#3073)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e38df45587bc50af9eec5f2e0aa262731ebc8a5d", "messageHeadline": "Add missing include to dnnl_conv_batchnorm.h", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "64feee1b52f54fef1dfaad39863b146459b41850", "messageHeadline": "Logging in framework.cc should use the session logger (#3059)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "fb871978b55ece9daced501859e215d184abd9f1", "messageHeadline": "Adjust build flags for the release pipelines (#3066)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 20, "deletions": 26}}, {"node": {"oid": "a7541f97530b4d0668a5147460dc55cee96f03aa", "messageHeadline": "[Java] Fix for incorrect input and output lengths in run call (#3064)", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 229, "deletions": 6}}, {"node": {"oid": "21cc2d88b4e97b343c8518ab893b45e886823dc3", "messageHeadline": "Move some vectors out of loops to minimize memory allocations/realloc\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 22, "deletions": 13}}, {"node": {"oid": "7ffb36be448bbf89fa3b0d52a577bf5e3f58c01d", "messageHeadline": "Fix softmax cpu code for double type (#3065)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "179603775f02176aca4d0a0bf00151c3ca228d26", "messageHeadline": "Use CUDA 10.1 for Linux build (#3057)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 41, "deletions": 377}}, {"node": {"oid": "be12fb3143347c6823ae25d1048f0e4bdd9d3d63", "messageHeadline": "include winml x86 binaries in the drop-signed-nuget artifact (#3058)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "752d74928c2d9a9432bc2d10b0a19338ca733637", "messageHeadline": "Improve the efficiency of SVMRegressor code (#3054)", "author": {"name": "Prabhat", "email": "prabhatr@usc.edu", "user": {"login": "prabhat00155"}}, "additions": 9, "deletions": 16}}, {"node": {"oid": "92b8a7a2be76de41e55b6026affc2af487f3e66f", "messageHeadline": "GPT2 Gelu Fusion & Test (#3009)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 837, "deletions": 9}}, {"node": {"oid": "932ecaea3490b174451156709d84b2ebdc771d06", "messageHeadline": "Some documentation updates. (#3060)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 37, "deletions": 5}}, {"node": {"oid": "c033cfee92dc42d7beb05747c63c93016b42c052", "messageHeadline": "Check python version (#356)", "author": {"name": "William Tambellini", "email": "william.tambellini@gmail.com", "user": {"login": "WilliamTambellini"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "45ba325fa659c22a8482287c71987384c6d6e833", "messageHeadline": "Remove USE_NSYNC macro (#3052)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 26, "deletions": 133}}, {"node": {"oid": "a1db87b3824a990fe4eeb4f71549ee4f02914d0d", "messageHeadline": "Add SafeInt bounds checking to memory allocation size calculations. (\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 273, "deletions": 115}}, {"node": {"oid": "21f9a8bdc2bfc4e78711e0846cc1e5ad330cfacc", "messageHeadline": "Allow creation of string tensor sequence (#3048)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 40, "deletions": 1}}, {"node": {"oid": "cb24e2a21438072e3496a0b8ed2ef0ee0e7e78b0", "messageHeadline": "Update nsync", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "85c0989e6fbaebc2d6a0898e0a0f0adbab1d6cee", "messageHeadline": "Fix build errors in benchmark", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 58, "deletions": 35}}, {"node": {"oid": "5306a1241b7f26141fec75c980ad9675c438b387", "messageHeadline": "Server: Add build options for the other EPs", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 37, "deletions": 0}}, {"node": {"oid": "44d5eaf3d7c1541ce8c91ee171c2b8959f28e831", "messageHeadline": "WinML exists in the nuget packages but does not publish its WinMD and\u2026", "author": {"name": "smk2007", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 36, "deletions": 0}}, {"node": {"oid": "bcbc5ee8155993358f89414e97b38e33f75a737a", "messageHeadline": "Improve error message for type mismatch between data in initializer a\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 18, "deletions": 17}}, {"node": {"oid": "dde4df148b9bf9d0059c5ae407f677144cf79d14", "messageHeadline": "Add and run spotless formatter for Java and Gradle", "author": {"name": "James Yuzawa", "email": "jtyuzawa@gmail.com", "user": {"login": "yuzawa-san"}}, "additions": 12566, "deletions": 10343}}, {"node": {"oid": "ef2bba316b74e2e47dcfe0f1dcd2efe46d500750", "messageHeadline": "CUDA 10.1 for Windows(#3049)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 10}}, {"node": {"oid": "18aa0940c20b5707a89123bb274f390e3018b6bc", "messageHeadline": "checking condition is wrong (#3025)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 0, "deletions": 7}}, {"node": {"oid": "e3c27536d08073015dd32737f5a7a79d18d1bf92", "messageHeadline": "Python binding doesn't need to link to the python lib on Linux", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "daf8c4bee40c45dc6b1b820be0fd14d45014ccfa", "messageHeadline": "Remove faturizers from CPU MLDNN and NoContribOps builds. (#3039)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "411b3aa801830fcaed5b49e0d788fbce780ef4e1", "messageHeadline": "Java build system enhancements (#2866)", "author": {"name": "James Yuzawa", "email": "jtyuzawa@gmail.com", "user": {"login": "yuzawa-san"}}, "additions": 412, "deletions": 298}}, {"node": {"oid": "ecdcd682bb347bce06ffc285ddf5f22000be8b10", "messageHeadline": "NCHWc ReorderOutput->Transpose(NHWC) fusion (#3035)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 534, "deletions": 195}}, {"node": {"oid": "71ca43b345e5a5cc52338826ba20cf3615ef3da3", "messageHeadline": "Revert an op version change (#3026)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 19, "deletions": 19}}, {"node": {"oid": "cc8adc87c3ae2aa600bae633ac7f62112198d84b", "messageHeadline": "fix NCHWc intermediates as graph outputs (#3033)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 36, "deletions": 4}}, {"node": {"oid": "4ca50d9352568db044597d7d44d3d45ff3f5633f", "messageHeadline": "Update DNNLibrary to v0.9.0 and update NNAPI GetSupportedNodes", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "aea76b078693db9c320f2beaac35415df93ac03d", "messageHeadline": "Bert optimization for onnx model exported by keras2onnx (#3014)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 845, "deletions": 389}}, {"node": {"oid": "3626c46fad21b688101e9614dedbaf793933493e", "messageHeadline": "Optimize Math::Im2col + ConvInteger pointwise (#3017)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 364, "deletions": 192}}, {"node": {"oid": "f98f0f87178c8848249eb2a77549243bb35480de", "messageHeadline": "fix type constraint name of Resize (#3019)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 25, "deletions": 25}}, {"node": {"oid": "2b77cb19bd306926693512820dac7d6850d8e925", "messageHeadline": "merge training kernels to master (#2999)", "author": {"name": "ytaous", "email": "4484531+ytaous@users.noreply.github.com", "user": {"login": "ytaous"}}, "additions": 2871, "deletions": 707}}, {"node": {"oid": "584ba714854c3b293436d5bbc538b24dc43d2120", "messageHeadline": "TensorRT dockerfile updates (#3016)", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "c9f18756b96d005497b6c962a8229788e17b3166", "messageHeadline": "Avoid unnecessary copies of ModelProto from being made in the Inferen\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 22, "deletions": 19}}, {"node": {"oid": "69bc8ce3c2caeab1fc816241d2c035f5e3af507d", "messageHeadline": "Upgrade protobuf to 3.11.3", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "5f29c9c3661ccf3e7a272231bde642a49c631e1a", "messageHeadline": "Add some test cases (#3010)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 464, "deletions": 191}}, {"node": {"oid": "c6b499bfc9f92a745b1ab75bc75b698b5b2b9af9", "messageHeadline": "Fix batch mode output for TfIdfVectorizer (#3008)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 39, "deletions": 36}}, {"node": {"oid": "da653ccdac344372052025d95cc84e93237f6f84", "messageHeadline": "Upgrade TensorRT to version 7.0.0.11 (#2973)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 452, "deletions": 176}}, {"node": {"oid": "273868eaa55431164a2e5b5a94b8941bd550a7c3", "messageHeadline": "Disable NuGetPackaging on Linux GPU and remove DML from the pipelines\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 5, "deletions": 30}}, {"node": {"oid": "ed0875909ed0037c538ab0bf04e057294efdba34", "messageHeadline": "Reorganize and update readme (#2916)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 98, "deletions": 107}}, {"node": {"oid": "e5f7e04e6ffb70541b589a0ffccd86b9f18ea271", "messageHeadline": "Use target_add_dml function (#3005)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "abb626ff60c9dabb064c08d8173a78266e0bec5c", "messageHeadline": "Provide alternative std::mutex implementation on Windows (#3000)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 137, "deletions": 44}}, {"node": {"oid": "c1997db85e87dad9e103bfe7035f4770ad68be3c", "messageHeadline": "Exclude faturizers from Linux NuGet packaging.", "author": {"name": "Dmitri Smirnov", "email": "dmitrism@microsoft.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "36915b36749b18e530dcf9d8b2c12381a018806c", "messageHeadline": "Temporarily remove Featirizers from packaging-pipelines", "author": {"name": "Dmitri Smirnov", "email": "dmitrism@microsoft.com", "user": {"login": "yuslepukhin"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "ce713823cc197fb7c49d654bb316041ffb3f356c", "messageHeadline": "enable winml in the gpu ci pipeline (#2993)", "author": {"name": "smk2007", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "5c5ac34b5c0ef0e1a5c7ad148a6a54ee486649be", "messageHeadline": "Disable use_dml in nuget pipeline (#3001)", "author": {"name": "smk2007", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fb2182f3fcd721c61d576533ed14b3715c7eef90", "messageHeadline": "Release ARM/ARM64 Nuget packages (#2987)", "author": {"name": "Tiago Koji Castro Shibata", "email": "ticastro@microsoft.com", "user": {"login": "tiagoshibata"}}, "additions": 102, "deletions": 42}}, {"node": {"oid": "3afb83ac3c66840b99c5854d0a697ebeb8f4ff94", "messageHeadline": "Support a few new C/C++ APIs (#2794)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 337, "deletions": 47}}, {"node": {"oid": "7437928f47eb716a37537f3946d5e0e30df3d5b7", "messageHeadline": "Replace hardcoded State serialization for Featurizer kernel tests (#2\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 122, "deletions": 140}}, {"node": {"oid": "64deb8030f789efb5974ebdedae92205784841dd", "messageHeadline": "Update ABI_Dev_Notes.md (#2959)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 10}}, {"node": {"oid": "b8d7d81506b93a09d0f94b2e5ecd3725d6144888", "messageHeadline": "Calling WinML enables telemetry by default for 1.2 (#2988)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 10, "deletions": 0}}, {"node": {"oid": "ca7626d7392b486af9950cad70112c4e3264e000", "messageHeadline": "Winml should be built against v2 of the capi as that includes the new\u2026", "author": {"name": "smk2007", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c8ea154e55e8eb772f9843b6abc16ba63d770847", "messageHeadline": "Package data_frame_tool, include featurizers into Manilinux2010 (#2989)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 17, "deletions": 8}}, {"node": {"oid": "62383b0328a4ecd91aba411ae072e753521ce0b3", "messageHeadline": "Improve BERT optimization script: Gelu and LayerNorm for models from \u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 555, "deletions": 109}}, {"node": {"oid": "0beb75ce772d11f71047ff622ca9315618361f3f", "messageHeadline": "populate file metadata for onnxruntime.dll (#2978)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 103, "deletions": 14}}, {"node": {"oid": "4f4f4bcd924f403618470045c158a02ed5443cdf", "messageHeadline": "Handle dynamic shapes and reshape the input according to the model (#\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 49, "deletions": 23}}, {"node": {"oid": "ec07fdc5b016d5e42b6e4bcd63159854d0aa4e4a", "messageHeadline": "Update ThirdPartyNotices.txt for eigen (#2985)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "bc9f55df47d6a8271735490d0deac432bf3d57ac", "messageHeadline": "Change eigen submodule url (#2975)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "8888b715624b8a85a07d2a5c69ea57f42285f849", "messageHeadline": "Add CounterVectorizer and TfIdfVectorizer. (#2981)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 427, "deletions": 0}}, {"node": {"oid": "0e5582bcc39a62a1628df4b9a1f023ec61809051", "messageHeadline": "Publish nightly NuGet packages to a new Vsts feed (#2976)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "f03e720727b0430f81fa611086bdd4ba1de8dcf1", "messageHeadline": "Disable _tlgPragmaUtf8Begin and _tlgPragmaUtf8End as well. (#2977)", "author": {"name": "smk2007", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 34, "deletions": 2}}, {"node": {"oid": "c32cedc6c9011f3e4069d686c6062fc7bfb55dcc", "messageHeadline": "Merge windowsai (winml layering) into master (#2956)", "author": {"name": "smk2007", "email": "smk2007@gmail.com", "user": {"login": "smk2007"}}, "additions": 31786, "deletions": 1165}}, {"node": {"oid": "c86db8067d9a37d11a1ef00ec416472cffcd411c", "messageHeadline": "Deliver a list of Featurizers (#2958)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2691, "deletions": 437}}, {"node": {"oid": "d5efbcb8d88fbe16d60d2cb39d09699d141fb847", "messageHeadline": "MLAS: Apply 'small-M' optimization for column-vectors (#2971)", "author": {"name": "Maxim Kalinin", "email": "makalini@microsoft.com", "user": {"login": "MaximKalininMS"}}, "additions": 29, "deletions": 0}}, {"node": {"oid": "7ff5c0e5a3d16b59c3e8cc6849c89d32e58b5b4e", "messageHeadline": "CMake changes (#2961)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 894, "deletions": 955}}, {"node": {"oid": "1d79926d273d01817ce93f001f36f417ab05f8a0", "messageHeadline": "Add int64_t support for split (#2944)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 22, "deletions": 10}}, {"node": {"oid": "79f1756d8ef7cc3ef630d00061bfe7c7b0bebcf8", "messageHeadline": "Removed build.py unused test_data_url and test_data_checksum options.", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "1e4080061bb6310fa24d7d02e2223a19b3dc186a", "messageHeadline": "Added support for double in batch norm (#2941)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 122, "deletions": 100}}, {"node": {"oid": "51595d6a4a8e8f65d888a98288ff68f2beac3025", "messageHeadline": "Add Int64 support to Neg", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "c23bd93ea5d2cda5d99fca3f311ac240b6316fed", "messageHeadline": "The closed variable in OrtEnvironment should be volatile to ensure th\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4cc0031177c211eae6069a84f3c5434e9626d9b3", "messageHeadline": "Java - Fixed a reference counting bug in the OrtEnvironment close met\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 13, "deletions": 1}}, {"node": {"oid": "7bb5c357a8cb9de21e31c554ef0de7728d4cc88f", "messageHeadline": "Adding some unicode to the String inference test to make sure it does\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "f33759883c20f5b616e6af2346dcf3a2b0679b6a", "messageHeadline": "keep output of bert model as float32 when converting to float16 (#2913)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 34, "deletions": 9}}, {"node": {"oid": "a319ad29b3b35bd2bde876ca31754f935d6ff85b", "messageHeadline": "Do not link onnxruntime jni output with JNI_LIBRARIES", "author": {"name": "Mina Asham", "email": "mina.asham@hotmail.com", "user": {"login": "mina-asham"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8a1de1a582bd3996f28b0ac917b6facac252eaa9", "messageHeadline": "Java - requesting JNI_VERSION_1_6 rather than JNI_VERSION_1_8 to supp\u2026", "author": {"name": "Adam Pocock", "email": "adam.pocock@oracle.com", "user": {"login": "Craigacp"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "f60badc1f34eed258220b609fdb1ff46f279f43d", "messageHeadline": "Documentation for API Breaking Changes (#2927)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 58, "deletions": 7}}, {"node": {"oid": "1239de3efec67db3de0974f86958973018743a3a", "messageHeadline": "Fix output shape computation in ConvTranspose op (#2688)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 30, "deletions": 11}}, {"node": {"oid": "e03e6676c0bd0acf7b80f840d9c9245b6389887b", "messageHeadline": "Add support for int64 to int64 LabelEncoder (#2926)", "author": {"name": "niklas-mohr-jdas", "email": "60099887+niklas-mohr-jdas@users.noreply.github.com", "user": {"login": "niklas-mohr-by"}}, "additions": 41, "deletions": 0}}, {"node": {"oid": "645d8fb2135c65fee32c7479ce98f26e0dc01df4", "messageHeadline": "Jignparm/upgrade macos vm image (#2928)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 8, "deletions": 71}}, {"node": {"oid": "653ce36ccf09355c77888c3a5593adba9eba9897", "messageHeadline": "Disable stacktrace on android", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "49ce4891bced02db4554253f98e595f1857b58f8", "messageHeadline": "Add noexecstack linker flag", "author": {"name": "James Yuzawa", "email": "jtyuzawa@gmail.com", "user": {"login": "yuzawa-san"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0279682147171662441250819460fa43ef1eff6b", "messageHeadline": "Add document for onnxruntime server.", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 24, "deletions": 20}}, {"node": {"oid": "546d8f71abb966f79f94f61ef04154b5f81e61ec", "messageHeadline": "Contributing: fix typos (#2905)", "author": {"name": "Maher Jendoubi", "email": "maher.jendoubi@gmail.com", "user": {"login": "MaherJendoubi"}}, "additions": 22, "deletions": 22}}, {"node": {"oid": "daff4240f0ea9a8a6fefd0d1bae0a75232d60843", "messageHeadline": "Updated README.md (#2910)", "author": {"name": "Saquib Nadeem Hashmi", "email": "saquibnadeemhashmi@gmail.com", "user": {"login": "Saqhas"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "cd876720d930fcea33c2c3f3bb9717ae115c007c", "messageHeadline": "Only fuse when output count of add is 1 (#2884)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 23, "deletions": 11}}, {"node": {"oid": "a92e924ab29ed71f2176bb04fd9ec3c117c0de5b", "messageHeadline": "Revert \"Use IArenaAllocator::Reserve for initializers and mem pattern\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 6, "deletions": 70}}, {"node": {"oid": "e0c9cdaa73e2de6da4d344c6366744589751982c", "messageHeadline": "Fix the nuget pipelines (#2901)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 82, "deletions": 63}}, {"node": {"oid": "17b72d55784f84718592ba56fe421112eadda522", "messageHeadline": "Fix NCHWc BatchNormalization regression (#2903)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 70, "deletions": 43}}, {"node": {"oid": "ba336b5583d4f8c44c4eb6735072b4d753617c03", "messageHeadline": "Disable DML EP on software adapter, fix float16 fallback bug, re-enab\u2026", "author": {"name": "Jeff", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 171, "deletions": 78}}, {"node": {"oid": "201b089a361f07420bab47f06ae16bc0971a0d50", "messageHeadline": "Fix some warnings on Windows (#2560)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 620, "deletions": 350}}, {"node": {"oid": "49725f896c78b42f539ca7557f008f033e444b36", "messageHeadline": "Disable openmp for the nocontribops pipeline. (#2888)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "fc51473b094bfcee521b13a0972651e7cb97b054", "messageHeadline": "Update BFCArena logic to use backoff if cudaMalloc fails. Makes behav\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 148, "deletions": 34}}, {"node": {"oid": "061f10fcd58c2c1eb6d4ccc8916db99ee21792e2", "messageHeadline": "Fixed typo in ORT_RETURN_IF_NOT() message. (#2862)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9f5e8c4ae81706e2080485c4ce5ac813ae95f34d", "messageHeadline": "InferenceSession::Run needs to call OnRunEnd for any EP that OnRunSta\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 24, "deletions": 4}}, {"node": {"oid": "38b34babe0e122d9c1db813bbef552fac36a6e8f", "messageHeadline": "Rashuai/boost cuda TopK performance (#2826)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 419, "deletions": 25}}, {"node": {"oid": "08113b80cc3c4bd0c446cd479d9a919f2a23d00d", "messageHeadline": "Optimize BatchNormalization to NCHWc Conv (#2855)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 180, "deletions": 12}}, {"node": {"oid": "807a59c55df347a6abf69737313b1e6b1b2ef50a", "messageHeadline": "Add calibration tool (#2845)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 561, "deletions": 35}}, {"node": {"oid": "22d9f3998e36c40bbf502f8333bf62de383b2cbe", "messageHeadline": "Fix positive raw scores for TreeEnsembleClassifier (#2824)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 2, "deletions": 9}}, {"node": {"oid": "b21576eeb0d8ca59b70ba39dee93eb6149f84195", "messageHeadline": "Support non-sequence tensor fed through as a python list (#2782)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 58, "deletions": 7}}, {"node": {"oid": "f9f25ec047019e222d7f77ca30213100d07dd215", "messageHeadline": "Fix spurious component detection warning (#2857)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 109, "deletions": 98}}, {"node": {"oid": "25d7ad187ffe701c22f551b86dd0d55a10008593", "messageHeadline": "Add float16 support back in the bert fusion script (#2870)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "95f3eb6aebd1f7ce5c897521081f6a9e66030b5e", "messageHeadline": "Bert fusion script for Tensorflow squad (#2858)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1791, "deletions": 1284}}, {"node": {"oid": "01f3a33c388c29a7854e646e68701038cd2af17c", "messageHeadline": "update protoc path to match protobuf version (#2865)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e6f7658adefab33527b1dd5e586b1eb7fbf326fd", "messageHeadline": "Update Windows GPU build to use cudnn 7.6", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "3853ddf9c733bc09b5b8bba7fed2cf3a254b7615", "messageHeadline": "Fix topk type handling to accommodate more types. (#2842)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 51, "deletions": 25}}, {"node": {"oid": "47e27ec9a14b6373d56f348c9ae14e96edecf7e0", "messageHeadline": "Disable DML in Windows GPU CI build (#2856)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "724ff0753bc3c7a1931e58ccb092df41b262d61f", "messageHeadline": "Use IArenaAllocator::Reserve for initializers and mem pattern planner\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 70, "deletions": 6}}, {"node": {"oid": "928b6bb21010b06c446248fa89ee95f11e61dded", "messageHeadline": "MLAS: enable threading for quantized GEMMs (#2844)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 602, "deletions": 166}}, {"node": {"oid": "5db8543018a2c2796defa714111633c32d4a2a33", "messageHeadline": "update optimization doc for BERT related fusions  (#2819)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 17, "deletions": 6}}, {"node": {"oid": "56030f8d74c55c05b3a5db918951939f9b32678b", "messageHeadline": "Fix Linux CUDA nuget packaging pipeline break", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "cff266e1b9f10b8af6cefa72a572fd219c3398cb", "messageHeadline": "Fix cgmanifest.json generating script (#2770)", "author": {"name": "Tiago Koji Castro Shibata", "email": "tiago.shibata@gmail.com", "user": {"login": "tiagoshibata"}}, "additions": 37, "deletions": 6}}, {"node": {"oid": "db05436fc02a8b07b9bfc5e72e9ea5273a484f5b", "messageHeadline": "User/orilevari/32bit comparison warning (#2800)", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "8643f3ebbb7d52952d4a356ea25b2e5375cedf51", "messageHeadline": "add domain check for nodes + update documentation (#2831)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 22, "deletions": 3}}, {"node": {"oid": "aa37dea5986280d1f12f4e3ecc12d44820401b63", "messageHeadline": "Convert ExternalProject Featurizers into git submodule (#2834)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 9, "deletions": 35}}, {"node": {"oid": "98cb41aa039b08c5d2451dd92124ab9a480f4099", "messageHeadline": "Ignore allocator type in ExecutionProviders allocator map. Make defau\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 155, "deletions": 121}}, {"node": {"oid": "b308e826a81040f0d817984a7c971bfff0f3e996", "messageHeadline": "Add support for int64_t for topk CPU. Fixes github issue #2806. (#2833)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 44, "deletions": 16}}, {"node": {"oid": "5c391854f468d97ec576b6b879a77a36cd8a4bec", "messageHeadline": "Upgrade gtest to the latest version (#2827)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 28, "deletions": 15}}, {"node": {"oid": "120433c29d9ec9de7f3603e9257b15cd5c7fa08a", "messageHeadline": "Add OneHotEncoder and HashOneHotEncoder kernels. (#2830)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 728, "deletions": 93}}, {"node": {"oid": "723cf8379300c5d054635135043d52c006d3883e", "messageHeadline": "Update Ubuntu & TensorRT version  in README (#2820)", "author": {"name": "Qing", "email": "cwq1913@gmail.com", "user": {"login": "Sanster"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "07502ec14eab0ab17807fd23be7c028be15dc519", "messageHeadline": "Fix dnnl wheel package name (#2823)", "author": {"name": "Yingge WAN", "email": "y-wan@users.noreply.github.com", "user": {"login": "y-wan"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "7c6242b0244731433b04d37d151271b322d3cb53", "messageHeadline": "update default optimization level + fix gemm_activation fusion (#2791)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 67, "deletions": 19}}, {"node": {"oid": "cc75e5a16243f031d6160d576417ae1999a4ab87", "messageHeadline": "update quantization doc (#2783)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 26, "deletions": 1}}, {"node": {"oid": "c4e4abce73965ff7819d3183712ce40e2ca13cb4", "messageHeadline": "Run static code analyzer on most of our code (#2817)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "e37cdbed74e237cdd467fdd5433eae9c5da85d2f", "messageHeadline": "Add manifest missing comma", "author": {"name": "Dmitri Smirnov", "email": "dmitrism@microsoft.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c4f6db7796dffef06d392948fdfe7cc7a977b3b3", "messageHeadline": "Fix memory leak in TRT (#2815)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "afa48b7e13ec2f706ec0569e24f57b30b9c604dd", "messageHeadline": "Add timeseries imputer transformer featurizer kernel (#2813)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 622, "deletions": 7}}, {"node": {"oid": "48e042868f67a241e878e3cbaad938de39902df4", "messageHeadline": "Update test data (#2356)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 168, "deletions": 174}}, {"node": {"oid": "31200ed92ca5c6e26e4615a4c161ef04a3868041", "messageHeadline": "speed up Windows TRT CI (#2811)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 106, "deletions": 16}}, {"node": {"oid": "b0019ac7fef58c40d731850c74403fe706e85c77", "messageHeadline": "add interface to copy batch tensors. (#2807)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 16, "deletions": 15}}, {"node": {"oid": "7ef6570e2715569dc907208855e702c7b8076484", "messageHeadline": "MLAS: update SGEMM threading parameters (#2808)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 0, "deletions": 13}}, {"node": {"oid": "71b5165ed39d13549b9a83194c283e2fc898b2a5", "messageHeadline": "Initialize max of softmax with lowest of float (#2786)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 23, "deletions": 23}}, {"node": {"oid": "2c8179bee44a53cc51ecd1c78adf43dcc2f19d3b", "messageHeadline": "ML.NET team needs featurizers within a package (#2789)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "1978376e1eff681d02bf69c10d8af2821b821295", "messageHeadline": "add session creation time cost. (#2798)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 14, "deletions": 7}}, {"node": {"oid": "32c5e76a16eba802c9faa689f275fd792905649d", "messageHeadline": "Improve bert optimization script: (#2712)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 101, "deletions": 39}}, {"node": {"oid": "f84240db2bfe314fc325269800d4ca0a6df3d5c0", "messageHeadline": "add uint8 support to where op (#2792)", "author": {"name": "Nathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "ebfcad1c90940a69fd370ae80110d4b1aa6fc949", "messageHeadline": "Add script for release Nuget validation (#2719)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 182, "deletions": 0}}, {"node": {"oid": "3e6f1836eb8e6e51602c07e2539b2df90505528f", "messageHeadline": "ACL EP convolution improvements (#2774)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 79, "deletions": 20}}, {"node": {"oid": "fdc0106f8309fc3061d54afa7caf1c38c0ef6d76", "messageHeadline": "ACL EP GEMM improvements (#2780)", "author": {"name": "Andrews548", "email": "32704142+Andrews548@users.noreply.github.com", "user": {"login": "Andrews548"}}, "additions": 20, "deletions": 38}}, {"node": {"oid": "f22bffe0f62460850efda82ee2cc38c93da6ca48", "messageHeadline": "Contributing: Fix a typo (#2784)", "author": {"name": "Maher Jendoubi", "email": "maher.jendoubi@gmail.com", "user": {"login": "MaherJendoubi"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "72bdfc8cd47175635cb3cc8f055503a2d9706abb", "messageHeadline": "Implement a more stable softmax (#2715)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 77, "deletions": 26}}, {"node": {"oid": "6f66260372bfa0e6f804eb7fe95c2d275a1442ec", "messageHeadline": "Import more featurizers (#2781)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1379, "deletions": 367}}, {"node": {"oid": "1b231180561cdfa9d0e2b74a511e050018b4ba58", "messageHeadline": "Fix nightly build version number issue", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "e3f674b5635ca360b111ab30eecc87602c3cfb66", "messageHeadline": "Disable featurizers in python packages", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "7ace7a5bcd0b65b91b1bb3462118343663633a98", "messageHeadline": "Pass BUILD_BUILDNUMBER to linux docker", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "7666d130e51873d1ff92f86670ba04ebcbf9496b", "messageHeadline": "Rename MKL-DNN to DNNL to fix broken link (#2730)", "author": {"name": "Changyoung Koh", "email": "gkcy1019@gmail.com", "user": {"login": "kcy1019"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "3d75c1b6be50b552e2734298d1355668a4ed3cd9", "messageHeadline": "Fix typo in error message (#2736)", "author": {"name": "Takeshi Watanabe", "email": "take-cheeze@users.noreply.github.com", "user": {"login": "take-cheeze"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "013642ed37b1bf78dbc17bbc9d8893a906914279", "messageHeadline": "Revert \"Change default optimization level to All (from Basic) (#2745)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 16, "deletions": 18}}, {"node": {"oid": "2096f324f436a8ba7861f9434935f68e37e104d6", "messageHeadline": "[NupharEP] fix a race condition when multiple sessions running differ\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 8, "deletions": 7}}, {"node": {"oid": "a2c8981a9e01aa87d2ee38a113ad6bbffd218bd3", "messageHeadline": "Add basic stacktrace output for posix debug builds. (#2749)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 36, "deletions": 2}}, {"node": {"oid": "382fa86af8703db4f5868d511632592d2058f568", "messageHeadline": "Pipeline changes for python 3.8 (#2753)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 28, "deletions": 8}}, {"node": {"oid": "fd334aff44b5e5dfe698716a31808c9d002586b3", "messageHeadline": "Update numpy to 1.18 (#2758)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 43}}, {"node": {"oid": "56bb503c2f26474b6613bcb2a198691a11dcef38", "messageHeadline": "Change default optimization level to All (from Basic) (#2745)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 18, "deletions": 16}}, {"node": {"oid": "90b708f8a9b8191593ee0d9dbe4e5f8037921749", "messageHeadline": "Update protobuf to 3.11.2 (#1928)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 182, "deletions": 664}}, {"node": {"oid": "b40a85a0e8f89928b965a5ed5e90e158cd642355", "messageHeadline": "Fix build for Python 3.8 (#2747)", "author": {"name": "Changyoung Koh", "email": "gkcy1019@gmail.com", "user": {"login": "kcy1019"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "c7a9c6b488c43990a532350cc788478e9585395d", "messageHeadline": "Split onnxruntime server to a separated folder (#2744)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1188, "deletions": 1802}}, {"node": {"oid": "6e4ec9b049d43bdb11c2ee45f622ea1e19ccacc6", "messageHeadline": "fix ngraph wheel (#2737)", "author": {"name": "zhanyi-ms", "email": "zhanyi@microsoft.com", "user": {"login": "mszhanyi"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9013fe6340574264d3cf33a430da0adc7a708c0b", "messageHeadline": "Fix build on Mac OS (#2731)", "author": {"name": "Andrey", "email": "andrey.lompart@gmail.com", "user": {"login": "tinchi"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "f142b683a129861cde6d73b120ee07a20df4773c", "messageHeadline": "Fix broken python docs links (#2740)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "ebf23744eb1e587e12e3459b1891a89a840fc687", "messageHeadline": "MLAS: ARM64 build fix (#2734)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "71ce0b8e0a35a2f762a397fbb4d54f02dc8d64de", "messageHeadline": "remove 64bit warning message from python validation. (#2727)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "abda4780839119668893a652caf78d65542d18cb", "messageHeadline": "[server] Add supposed for model_name and model_version as cli paramet\u2026", "author": {"name": "Andrey", "email": "andrey.lompart@gmail.com", "user": {"login": "tinchi"}}, "additions": 16, "deletions": 4}}, {"node": {"oid": "cd6ff42442b43de90785bcf0d37df575273c4959", "messageHeadline": "Add schema for new Qops (#2611)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 421, "deletions": 1}}, {"node": {"oid": "71940c09157399beabec57b545d81047b369bbd0", "messageHeadline": "Update Nuphar tutorial notebook (#2721)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 280, "deletions": 165}}, {"node": {"oid": "bbe32f0990c4bbc8b0dc5a2221ee24b705e2a4ee", "messageHeadline": "Extend DML kernels (#2641)", "author": {"name": "Jeff", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 621, "deletions": 211}}, {"node": {"oid": "4b900dc585045601f04326bb0ea5e4b565cd6241", "messageHeadline": "Simplify cache implementation and avoid static variables that may car\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 121, "deletions": 127}}, {"node": {"oid": "da03ed447350314f662ee06d7829f2591ff25bef", "messageHeadline": "Tiny fix to codegen", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 9}}, {"node": {"oid": "b42cb619044df502a90fcbde1a326948719e0909", "messageHeadline": "Packaging pipeline changes for VS 2019 (#2711)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 225, "deletions": 551}}, {"node": {"oid": "f0046a153113732e802d55e17e101fa1c727d9ce", "messageHeadline": "switch back order of session_state_ and execution_providers_ in Infer\u2026", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "7c0235c15a835e7c9c31c75509e6af191d7ec5c6", "messageHeadline": "Propagate documentation modification from rel-1.0.0 (#2713)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 10, "deletions": 27}}, {"node": {"oid": "967ffc03cc32c7867edd5d9633f382a878e119b3", "messageHeadline": "a subset of the models in testdata were missing an ir_version. update\u2026", "author": {"name": "Ori Levari", "email": "ori.levari@microsoft.com", "user": {"login": "orilevari"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "72286d91e7f089fa9f8d85aa48c8a57e2e8b610f", "messageHeadline": "fix the CUDNN_BN_MIN_EPSILON difference issue between cudnn7.3 and cu\u2026", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "64112db346d905ed11b7dd680283d41b830e95fb", "messageHeadline": "Fix C# handling of unicode strings (#2697)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 63, "deletions": 37}}, {"node": {"oid": "233bdd268b861774efbe20e7b02a25ee48221f55", "messageHeadline": "Advance Featurizers commits and fix up ORT lining issues. (#2705)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 524, "deletions": 572}}, {"node": {"oid": "9b14d8fd6743d3424e454ceb8b0aafdcb1382396", "messageHeadline": "Update rnn_benchmark to reset OMP threads (#2698)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 47, "deletions": 44}}, {"node": {"oid": "bb7f43ee912588d6e0c01152ac6fc07b7ec33de9", "messageHeadline": "Documentation update: build instructions (#2636)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 73, "deletions": 67}}, {"node": {"oid": "b3d0b114fe9f3280218dcc624cb968fa586b015c", "messageHeadline": "Support Round op in the CUDA EP (#2601)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 44, "deletions": 11}}, {"node": {"oid": "715e3657239bc6b5d92f434e71b935677dfcfe27", "messageHeadline": "Enable int8 GEMV tensorization (#2696)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 37, "deletions": 202}}, {"node": {"oid": "971bc439b5a57b38c82c3e855f341676cbe4ede4", "messageHeadline": "Support CumSum op in the CUDA EP (#2647)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 459, "deletions": 18}}, {"node": {"oid": "9017e93701dc60af14ba2e667c17e2ac31348bd4", "messageHeadline": "[NupharEP] fix for Windows build and VS 2019 (#2694)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 11, "deletions": 1}}, {"node": {"oid": "89d6bfaa9462d3f494bfa04749767b88f34b0e90", "messageHeadline": "VS 2019 build pipeline changes (#2693)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 435, "deletions": 67}}, {"node": {"oid": "0d5504877b1665afb6c2352a11760667d6831c6e", "messageHeadline": "fixed \"unreachable code\" warnings for Windows build (#2682)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 12, "deletions": 13}}, {"node": {"oid": "504ba6f2ca4ec79f4c6c986d6bad2b6667b8436f", "messageHeadline": "Remove a warning", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "ce7a180f21fb97c531b073f3590cc928e7bf970b", "messageHeadline": "Import more featurizers with tests (#2685)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1292, "deletions": 426}}, {"node": {"oid": "c767e264c52c3bac2c319b630d37f541f4d2a677", "messageHeadline": "[NupharEP] update tutorial with GPT-2 (#2677)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 142, "deletions": 34}}, {"node": {"oid": "b38b36a94148d7dc025c3e094ccc02473e4bc2d5", "messageHeadline": "Increase max_num_graph_transformation_steps from 5 to 10 (#2633)", "author": {"name": "Jeff", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "9dddabf237bc2926c0008510885e7412af461f55", "messageHeadline": "Miscellaneous DML EP fixes (#2634)", "author": {"name": "Jeff", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 40, "deletions": 16}}, {"node": {"oid": "c9078818527c61f39ddb64468ed5bfab21fcbbc2", "messageHeadline": "MLAS: optimize QuantizeLinear (#2660)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 364, "deletions": 25}}, {"node": {"oid": "47503ec7a62975d502bb0ad4e589185029f84327", "messageHeadline": "Initiate the build scripts for ARM ACL (#2652)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 222, "deletions": 4}}, {"node": {"oid": "f7412899a107535d18fc63e464f14bb6987843eb", "messageHeadline": "added cache version for nuphar JIT binaries (#2646)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 123, "deletions": 23}}, {"node": {"oid": "7c87070b249987a790406609c0c7a2ca20d48cf8", "messageHeadline": "Import Featurizers (#2643)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 529, "deletions": 1305}}, {"node": {"oid": "a46a28b7d81634ecc4811177a28cdc83b90b48cc", "messageHeadline": "Windows CI changes(#2650)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 546, "deletions": 120}}, {"node": {"oid": "18bdde34dcc143dccacaeaa4c18a7d39dc9c72fb", "messageHeadline": "Fix memory exception in Layer Norm Fusion (#2644)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "1996129ddf59584bef6af5ceea1c6123468c4d3d", "messageHeadline": "Improve performance of resize() in Nearest mode (#2626)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 228, "deletions": 60}}, {"node": {"oid": "6e68007d2d029aa17282033a7c7008b10e8c07ca", "messageHeadline": "fix build error for ARM (#2648)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "fff1ed9bfc28b6e1cbb1c29364b8872e2929a5df", "messageHeadline": "Optimize cuda scatter() on 2D compatible. (#2628)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 184, "deletions": 64}}, {"node": {"oid": "e31be23c4225f0f4a65403d521e0187053446cd7", "messageHeadline": "Cuda pad optimize when no padding is needed. (#2625)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "8631b70c7355534138273303d380eec2ca0890f2", "messageHeadline": "Improve cuda expand() opeator's performance. (#2624)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 209, "deletions": 113}}, {"node": {"oid": "ac08b58867625b6314748bdbf825c59cc371e2fd", "messageHeadline": "Add shape inference to ConvTransposeWithDynamicPads schema (#2632)", "author": {"name": "Jeff", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 137, "deletions": 3}}, {"node": {"oid": "4dbf9442cc3469958092103a72d4f3a1ee16a523", "messageHeadline": "removed unnecessary batch file and fix path (#2640)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 3}}, {"node": {"oid": "c7cd336917fee8fcf9d8fb17088ce34276cae20d", "messageHeadline": "epsilon attribute for layernormalization fusion (#2639)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "d6f33dceb15fcbd1ad9cee1d47a326fab1fdc4a0", "messageHeadline": "fix float16 comparison in initializer (#2629)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c04647b952b82137693540efac220a3e33062200", "messageHeadline": "Improve Embed Layer Norm Fusion for SQuAD with static input shape  (#\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 453, "deletions": 208}}, {"node": {"oid": "6859d92e2bb2a4767ea7798bba91c303bf7dd21a", "messageHeadline": "Make sure fenced tensor could not reuse other tensor. (#2561)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b2d65b447f764928499f3c86999fff23d6e68ba2", "messageHeadline": "Remove unnecessary parameter in some places in GatherElements impleme\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 0, "deletions": 7}}, {"node": {"oid": "87297846352a148c7b08b08a9998098a98ebcb20", "messageHeadline": "Allow providers to be set for InferenceSession at construction (#2606)", "author": {"name": "Eric Cousineau", "email": "eric.cousineau@tri.global", "user": {"login": "EricCousineau-TRI"}}, "additions": 11, "deletions": 2}}, {"node": {"oid": "1ee250def883d2630c8ab18019509668dbdda6e3", "messageHeadline": "EmbedLayerNormalization Fusion For Dynamic Squad Model Opset 10 (#2613)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 147, "deletions": 37}}, {"node": {"oid": "2ca9733cee388e7cdf90d27728c8721832ffbac2", "messageHeadline": "Dump subgraph ID and fused graph ID (#2607)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 20, "deletions": 6}}, {"node": {"oid": "45babd6c005b6c8afdcf48456a9d20af8dd9d530", "messageHeadline": "symbolic shape inference: fix warnings in GPT-2 model (#2608)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 10, "deletions": 5}}, {"node": {"oid": "bc89eccb21e4674a753cee15820c43d5e35b5341", "messageHeadline": "Support opset 11 subgraph of Squad model in Embed Layer Normalization\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 306, "deletions": 76}}, {"node": {"oid": "796948c6aee80df5562c3e1538bb675913946754", "messageHeadline": "Rename automl python tools folder to featurizer_ops. (#2593)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "6858f0a06ab8f833de7171dba2f4c86319fb159c", "messageHeadline": "Add support for opset 11 in reshape fusion (#2592)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "35ceb1a6a66fcb5c485ea720169e9a085bfab7d1", "messageHeadline": "Java API for onnxruntime (#2215)", "author": {"name": "Adam Pocock", "email": "craigacp@gmail.com", "user": {"login": "Craigacp"}}, "additions": 41793, "deletions": 4}}, {"node": {"oid": "b0128a4843da4ddbf2d029248ec7d63a2d0ce05c", "messageHeadline": "Fixed an issue in updating realized dims (#2597)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 50, "deletions": 5}}, {"node": {"oid": "78099701b48549854e22e8d080aef2752c89478e", "messageHeadline": "Add missig env variables for mac pipeline test (#2595)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "41fc820f761ff92e6a1f1f19b0bf2065035ef81f", "messageHeadline": "add path to build dir before test run (#2590)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "6e08efa6a2c77403c9cb65bb3dc157eced419bbd", "messageHeadline": "Fix lto bug for protobuf and ubuntu", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "62de8fa8417b8c379291562cd1945cfd0c887c21", "messageHeadline": "Update docs for Android NNAPI EP (#2586)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 75, "deletions": 12}}, {"node": {"oid": "0ab54521f44a682a4c71a9f0415481adde84b4a8", "messageHeadline": "Temporarily exclude vgg19 test from Python backend test", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "36eb1771bab932043c03544b4b927725f738e9aa", "messageHeadline": "Update version (#2584)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "200f4b4ea616ef4a35aeced6795e997ecfa2349d", "messageHeadline": "EmbedLayerNormalization Fusion Improvement (#2553)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 154, "deletions": 56}}, {"node": {"oid": "0f12346d769ff5c603dccc4bcf3a4ba0d1820be4", "messageHeadline": "[Nuphar EP] fixes for some object detection models (#2581)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 64, "deletions": 37}}, {"node": {"oid": "cbc398bb759737bbb3f09e0c6050d6bcdccc6e90", "messageHeadline": "Ryanunderhill/packagename test (#2582)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "c06dbd83113f706d6c9769f3eb153a0eff737ead", "messageHeadline": "Add ConvTranspose1D (#2578)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 178, "deletions": 108}}, {"node": {"oid": "79847f39b3fa9fa40ac833a2e0a4646464e0ae44", "messageHeadline": "Fix file not found error during docker build. (#2569)", "author": {"name": "Mark", "email": "Exlsunshine@users.noreply.github.com", "user": {"login": "Exlsunshine"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "5575766a53586dc8c9a01603b0bef1426369ed96", "messageHeadline": "Add more check on SkipLayerNorm and BiasGelu fusion (#2574)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 31, "deletions": 11}}, {"node": {"oid": "262ee9dc5a7e33337206250587ef2a2245acb708", "messageHeadline": "Fix a warning found in the latest VS release", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 3}}, {"node": {"oid": "34beafc51ce37d61737e6d3bc932497aba35acfe", "messageHeadline": "make layernorm fusion to support opset 11 (#2545)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "eeb28a80c0d1e81421d3dee2242efadef255ba45", "messageHeadline": "setup java ci mac (#2570)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 8, "deletions": 11}}, {"node": {"oid": "038ee91da52e9167dd35a487cf289893a3b2ee0f", "messageHeadline": "Allow sequence length to be symbolic (#2559)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2, "deletions": 6}}, {"node": {"oid": "73c682b97c5f68b581eabfc4e3e9a2d5a8a3a1da", "messageHeadline": "disable onnx_test_runner -x invocations for dnnl (#2568)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "7eddac16c25570cbe4f9d48f1ec9dce34e14e86d", "messageHeadline": "Re-enable Windows C# tests (#2564)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "854362cf05b4c7c2053e38437a5e29c6f6a0f79e", "messageHeadline": "Update win-x86-ci.yml (#2557)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ace132f9aaad6d14c37223aa1ec0c78673d52eb0", "messageHeadline": "Fix android build (#2558)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "4c996a86997aac4ecaa051f3564c187817452433", "messageHeadline": "DNNL CMAKE update (#2548)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 1, "deletions": 44}}, {"node": {"oid": "53a6bc2f07d956d56f2c43ec714a76af4f311c95", "messageHeadline": "Fix a bug handling negative begin pad values in Pad op (#2550)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 52, "deletions": 4}}, {"node": {"oid": "bec4abf074f6b58e50d79d6a9de0f0751d14baa5", "messageHeadline": "Add back executable bit to build.py", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "281933fa1c487f29817cc5f05978319b8ffd4ab7", "messageHeadline": "Fix C API tests for centos and mac (#2544)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 32, "deletions": 12}}, {"node": {"oid": "d34fb62012d1926041ac5284b161ce2d737d3d22", "messageHeadline": "Introduce container type runtime checks and other improvements (#2522)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 996, "deletions": 661}}, {"node": {"oid": "be56d77a66dc807620253fc472465266071051c4", "messageHeadline": "Fix integer overflow in cuda NonMaxSuppression implementation (#2540)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 31, "deletions": 18}}, {"node": {"oid": "3e7aaf8fa13cffa60759de536cd10204bde7d3f3", "messageHeadline": "User/xianz/telemetry (#2458)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 110, "deletions": 15}}, {"node": {"oid": "293b15480be468529cbd05e0c150fad23b08f186", "messageHeadline": "Add dynamic shape support in TensorRT execution provider (#2450)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 412, "deletions": 146}}, {"node": {"oid": "d748f891d818a6bbde089cd88040582bd1bee300", "messageHeadline": "Revert \"Disable thread pool creation when enabled OpenMP (#2485)\" (#2\u2026", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 3, "deletions": 15}}, {"node": {"oid": "5c2e47475172cc3fd87fca55af8c43dc7dc95bfe", "messageHeadline": "Add provision in ORT for session options to be parsed when available \u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1103, "deletions": 195}}, {"node": {"oid": "178d059111f9c417cf5cb123a37ade9826b6c2fb", "messageHeadline": "Setup java ci (#2528)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 13, "deletions": 2}}, {"node": {"oid": "b50878dcf0534ec4031924e18d5a6043d8a363a3", "messageHeadline": "Disable Attention fusion tests when DISABLE_CONTRIB_OPS is defined (#\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "e32eff826ccb533dee444e837727a1ad2ec5f546", "messageHeadline": "enable nuget package testing on centos7 (#2527)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 43, "deletions": 6}}, {"node": {"oid": "85a4ed8cf77ef54edc4421ffb119779a09106fc7", "messageHeadline": "fix cuda kernel causing invalid mem access (#2523)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "66254eb25a9e5985b78ef33449c1447c73d07959", "messageHeadline": "Update BERT model optimization python script (#2521)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 546, "deletions": 166}}, {"node": {"oid": "31ea11a696cc8590f3260b99503d7bddc3a73d56", "messageHeadline": "Renaming MKL-DNN as DNNL (#2515)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 1209, "deletions": 1203}}, {"node": {"oid": "3d627362a0c4e27b90b237c573237e375b0ce154", "messageHeadline": "Upgrade Windows CPU CI pipeline to use VS 2019 (#2519)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 250, "deletions": 20}}, {"node": {"oid": "e8b327d6573845654700c048c330c9c722f3d3f8", "messageHeadline": "Fix constant folding of node assigned to CUDA (#2510)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 79, "deletions": 36}}, {"node": {"oid": "4354023913a9f84a04a5dec25ce80a5fd92e1e78", "messageHeadline": "Make link time optimization work on Linux (#2477)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 10}}, {"node": {"oid": "25c260fdef28a955919fe0d375a5aa3cf19f229c", "messageHeadline": "Add parallel for tensorized gemm (#2517)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 18, "deletions": 2}}, {"node": {"oid": "c1be615c4538e57c5878d0973e0bb44d043c2046", "messageHeadline": "[NupharEP] refine parallel schedule control (#2514)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 84, "deletions": 31}}, {"node": {"oid": "784eca0dcdf6ff24e3e55f9166d6e12d5617dd17", "messageHeadline": "Cuda pad() for opset 11 (#2490)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 86, "deletions": 10}}, {"node": {"oid": "b9faa0b6fdb8fe61b076c2ac342c98ebe5597e0b", "messageHeadline": "Fix kernel registry validation to reenable DML kernels", "author": {"name": "Jeff Bloomfield", "email": "jeffbloo@ntdev.microsoft.com", "user": {"login": "jeffbloo"}}, "additions": 25, "deletions": 9}}, {"node": {"oid": "ddaad866058d0792aaee83018d3b58a6ded958d4", "messageHeadline": "CUDA Loop (#2444)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 276, "deletions": 101}}, {"node": {"oid": "50eb14011907adefcfa88ba2ce26aaf6c591f4cf", "messageHeadline": "Cuda Resize Operator for opset 11. (#2484)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 692, "deletions": 302}}, {"node": {"oid": "c42148a0c37162a2d211d8ebd125bc560272740d", "messageHeadline": "Improves softmax function for standard ml", "author": {"name": "xavier dupr\u00e9", "email": "xavier.dupre@gmail.com", "user": {"login": "sdpython"}}, "additions": 11, "deletions": 17}}, {"node": {"oid": "ec88f6d8d604ddde7db6be83ede3b270f16b0dc5", "messageHeadline": "Add DataFrameTool (#2456)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 291, "deletions": 0}}, {"node": {"oid": "89824b35e950190da1ebeb9e3c8a0bc9ace0c86f", "messageHeadline": "optimize CPU implementation of Attention (#2496)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 148, "deletions": 123}}, {"node": {"oid": "0f57e0a49e98109e8e11d97b8362988d81d6ee2c", "messageHeadline": "Change mask input of EmbedLayerNormalization op to be optional (#2495)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 188, "deletions": 124}}, {"node": {"oid": "0edd4ef6ca7323659693eb92b3c9b3a97f13970d", "messageHeadline": "EmbedLayerNormalization fusion (#2452)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 341, "deletions": 5}}, {"node": {"oid": "60208463a9bb0a96bbdd693a89d405a04c8e3235", "messageHeadline": "[NupharEP] Enable parallel schedule (#2505)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 170, "deletions": 32}}, {"node": {"oid": "005305be6e201c7ac0d8619f6813027165c521e5", "messageHeadline": "Implement AddGelu and SkipLayerNorm (#2487)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 401, "deletions": 206}}, {"node": {"oid": "ee0bde6b69154a18e8719fa6bf4324b0f38f1f38", "messageHeadline": "Enable three type of Equal() to version 11. (#2508)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 19, "deletions": 7}}, {"node": {"oid": "75b4747701b818a37af0dddca0d175f854e15964", "messageHeadline": "Fix a memleak in pybind. (#2503)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1fdf1006ac312295d164b2cbc0f88698122a7b8f", "messageHeadline": "Various fixes coming out of discussions in #2436 (#2497)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 29, "deletions": 10}}, {"node": {"oid": "04b6097db45610ddfa2c534ff0ea04589a1cc6ef", "messageHeadline": "Cuda Clip() for op set 11. (#2411)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 61, "deletions": 29}}, {"node": {"oid": "ccbd778d0d15f2597f2cc5d644a58e99b44cca27", "messageHeadline": "optimize CPU implementation of EmbedLayerNorm (#2491)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 56, "deletions": 30}}, {"node": {"oid": "e57b735bb9a4912eb38f9de9d272d42c03779569", "messageHeadline": "Add a transformer to use Gelu approximation for cuda provider (#2480)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 387, "deletions": 18}}, {"node": {"oid": "7c7d5a149c9ed52eec67304bae5c4b132166a8a1", "messageHeadline": "Disable thread pool creation when enabled OpenMP (#2485)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 16, "deletions": 3}}, {"node": {"oid": "e29fb5cef1d0e8ad27a4ede6f37606a368f3682d", "messageHeadline": "Add BatchParallelFor, TryParallelFor, TryBatchParallelFor into Thread\u2026", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 177, "deletions": 34}}, {"node": {"oid": "d6c84925d514a94d75ade49fb71a5f6f77885b0a", "messageHeadline": "move logSessionCreation after session is initialized (#2481)", "author": {"name": "Xiang Zhang", "email": "xianz@microsoft.com", "user": {"login": "zhangxiang1993"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "ca8ff8c91c3ae6f7587df3cea3bd82b5578613a9", "messageHeadline": "Enable conv/conv_transpose for opset 11 in cuda execution provider. (\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 84, "deletions": 32}}, {"node": {"oid": "1ca0e0866e905eb36046b6bd55dba2c6f1df4625", "messageHeadline": "Revert \"Remove DeviceAllocatorRegistry class (#2451)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 26, "deletions": 0}}, {"node": {"oid": "4cac18f6661f7e491297c5e3db7f1758da8ea4b4", "messageHeadline": "Optimize CPU Transpose for a single axis moving (#2461)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 457, "deletions": 10}}, {"node": {"oid": "882f28a74b1d7e156b896b081a5ca367db027222", "messageHeadline": "Fix NuGet end to end tests for custom op dll (#2472)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 47, "deletions": 3}}, {"node": {"oid": "5dce9be4f9d78e76afd52376332550a34d922f7d", "messageHeadline": "Add Attention Fusion Transformer (#2445)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1025, "deletions": 40}}, {"node": {"oid": "4caf5c9c136756cdf0877b873e983d3fcc22f301", "messageHeadline": "add additional test data set for nuget pipeline (#2448)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 202, "deletions": 51}}, {"node": {"oid": "e0b51e4d95fc26420a8a3cc30faf3873a6f6ad60", "messageHeadline": "Symbolic shape inference improvements: (#2460)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 262, "deletions": 106}}, {"node": {"oid": "a01816335d71ef9b144a403db4dd8aa997474446", "messageHeadline": "Allow more than one invocation of CreateEnv in the same process. (#2467)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 110, "deletions": 51}}, {"node": {"oid": "b3bfc094b92aecb311d4311f9debe7772a82dd44", "messageHeadline": "Fuse SkipLayerNorm with Bias (#2453)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 397, "deletions": 92}}, {"node": {"oid": "e7131f6d12b7eb8902fe15914f7b56d04a0fcb4a", "messageHeadline": "Pull the latest image before running docker build", "author": {"name": "Changming Sun", "email": "me@sunchangming.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 22}}, {"node": {"oid": "0341ee90600b605a78026bb38e9d6d80ebc33080", "messageHeadline": "Clean up build.py (#2446)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 51, "deletions": 114}}, {"node": {"oid": "d49cbf6e0820aa09deb5d17bf87597eb6309f984", "messageHeadline": "Parallel Gelu with ParallelFor (#2399)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 26, "deletions": 0}}, {"node": {"oid": "ca0ed966212ca3e011bb77a82ab1914d74605c7f", "messageHeadline": "CSharp api and test for loading custom op shared library (#2420)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 345, "deletions": 13}}, {"node": {"oid": "6a88d03621e61e68ead16c2ec5844e00cba383ed", "messageHeadline": "Remove DeviceAllocatorRegistry class (#2451)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 26}}, {"node": {"oid": "d486481455a1f2bb710d79a2de0c2e80552c3750", "messageHeadline": "Correctly handle implicit inputs for fused nodes (#2390)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 110, "deletions": 50}}, {"node": {"oid": "6bcf95477eca78e283a1282f545bcc4c34e4e82c", "messageHeadline": "Fix Windows GPU C API packaging pipeline failure (#2440)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 46, "deletions": 0}}, {"node": {"oid": "87d3794ad28d230a2608b6ca955145c6cd003510", "messageHeadline": "Add CUDA Scan operator. (#2403)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 263, "deletions": 47}}, {"node": {"oid": "9b7b5e2c27c54aaff0ea16f0662c9864ffc2405c", "messageHeadline": "Adjust codegen vectorization width from target (#2439)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 114, "deletions": 74}}, {"node": {"oid": "a5daa6faea638555071e96336ab37e635b24a6b4", "messageHeadline": "add Im2col<uint8_t> (#2438)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 197, "deletions": 115}}, {"node": {"oid": "3be554c2fbde8eb2951b231c6067d69325029668", "messageHeadline": "Fix the type constraints on CUDA If operator to exclude strings. (#2431)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "8647201ac739506cdc692439ed8eb73bc259daea", "messageHeadline": "Improved documentation for onnxruntime::utils::SwapByteOrderCopy(), a\u2026", "author": {"name": "Edward Chen", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 12, "deletions": 5}}, {"node": {"oid": "be12cdc73fabea7b50396b457bd0f2e1c92a6a96", "messageHeadline": "Add CUDA If operator. (#2377)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 106, "deletions": 15}}, {"node": {"oid": "1cb6bdc33c64c321a17b62c25c04bc9878d35ba9", "messageHeadline": "Added support for Pad-2 operator in OpenVINO-EP (#2405)", "author": {"name": "Patrick Foley", "email": "patrick.foley@intel.com", "user": {"login": "psfoley"}}, "additions": 35, "deletions": 0}}, {"node": {"oid": "95e8c3377e2b8e9b7ca1e8221e800a931ce42591", "messageHeadline": "onnxrt server documentation update (#2396)", "author": {"name": "avidiyal", "email": "41489828+avidiyal@users.noreply.github.com", "user": {"login": "avidiyal"}}, "additions": 19, "deletions": 12}}, {"node": {"oid": "0e947bd328a23b54a88487e08297c801cb91c912", "messageHeadline": "feat(treeregressor): Update TreeEnsembleRegressor for type support (#\u2026", "author": {"name": "Nick Groszewski", "email": "groszewn@gmail.com", "user": {"login": "groszewn"}}, "additions": 31, "deletions": 11}}, {"node": {"oid": "367361fc74c8c749f1b857e79ced15f7d94b5f1a", "messageHeadline": "Fix the issue in matmul_add_fusion (#2407)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 21, "deletions": 9}}, {"node": {"oid": "aa7c79eac9875ba3b5c9444b13d1878560415ea7", "messageHeadline": "[NupharEP] Update notebook and docker image (#2416)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 207, "deletions": 77}}, {"node": {"oid": "e1c17fd126a474b1e21a7c160eaaab334e82a545", "messageHeadline": "Add Reshape Fusion (#2395)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 578, "deletions": 53}}, {"node": {"oid": "f268e69c7928df40174038bc166fda6e8a5ca591", "messageHeadline": "Minor optimization: if a node has already been placed, there's no nee\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "5ab7041fa760c80063cb4a463530ac091b588845", "messageHeadline": "fix cross compile bug (#2415)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1e03ce84eb7a2148a90908e3510538f2a48f01be", "messageHeadline": "[NupharEP] force some low/zero cost ops to be inlined (#2409)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 36, "deletions": 3}}, {"node": {"oid": "c1d757a00b0a0f08ad2d1fb743477b7d1094d8d8", "messageHeadline": "Add opset 11 versions of the existing CUDA operators that had negativ\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 412, "deletions": 175}}, {"node": {"oid": "6cc57721f4cf6e10cb7f829ca66cf7aaa1a70ef4", "messageHeadline": "Change CUDA implementation of Transpose to support all fixed size ten\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 201, "deletions": 103}}, {"node": {"oid": "109b3cb4505095ee51d3008a89e1b2dcc01ca209", "messageHeadline": "Avoid using the default logger in the graph lib and optimizers (#2361)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 614, "deletions": 556}}, {"node": {"oid": "b15e43a541ed1e888d989bbbf894587ac8cfda64", "messageHeadline": "[NupharEP] Multiple optimizations  (#2380)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 758, "deletions": 324}}, {"node": {"oid": "7e164eaa6a26d82df0fff8ff9260c05ba850bbe3", "messageHeadline": "Fix reuse logic in allocation planner. (#2393)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 133, "deletions": 14}}, {"node": {"oid": "b90d55b7ea9e1e86624827145c85fbc3db9eec5b", "messageHeadline": "Fixed compilation with ngraph (#2388)", "author": {"name": "Ilya Lavrenov", "email": "sandye64@icloud.com", "user": null}, "additions": 5, "deletions": 4}}, {"node": {"oid": "dde410e073a948146256b625a7528211641aee89", "messageHeadline": "fix BUILD.md typo (#2375)", "author": {"name": "nihui", "email": "shuizhuyuanluo@126.com", "user": {"login": "nihui"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "51571030ef264aca9afcebf73243d53567857330", "messageHeadline": "Another try to stabilize CUDA CI (#2383)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 17, "deletions": 24}}, {"node": {"oid": "ffa28125872086c7429db8e4f72b4cf1a32da336", "messageHeadline": "Skip layer norm transform (#2350)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 159, "deletions": 1}}, {"node": {"oid": "8ed2928dd5ea9779549d66faf6286e16126a0773", "messageHeadline": "Fuse Add + Gelu (#2360)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 519, "deletions": 14}}, {"node": {"oid": "4b72fedbd51821535231cccab6bd8f03ca351aa1", "messageHeadline": "Layer Norm Fusion Fix (#2379)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 44, "deletions": 3}}, {"node": {"oid": "8c733c8d82fee80336f80085dbc8d29abbfabae6", "messageHeadline": "Add opset 11 version of Split to CUDA ops (#2376)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 76, "deletions": 65}}, {"node": {"oid": "c0d23d5ffe5a684d9972336724ff8c6e79cdd743", "messageHeadline": "Fix bug with Slice. Need to pass in flattened input dimensions so the\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 42, "deletions": 14}}, {"node": {"oid": "9e26f4de6fbce630e90e3d33507ab039d545b196", "messageHeadline": "Extend OneHot CPU kernel to support more types (#2311)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 8, "deletions": 0}}, {"node": {"oid": "437772d5bc492801ae6f068877bfd9116f3d36c2", "messageHeadline": "update output size calculation for resize (#2366)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 276, "deletions": 78}}, {"node": {"oid": "192dcfaa8ea4ad1b2258ae88a02f8dbc6c4b971c", "messageHeadline": "Fix a bug in TLS refcount that may destabilized CUDA CI (#2374)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 0}}, {"node": {"oid": "41b9f01e4c935eef3662a6a58a7ab7fd0de13041", "messageHeadline": "test bidaf with nuphar for avx target (#2370)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "fc6773a65b9d86e5261d108c157846ddda76385b", "messageHeadline": "Add Tracelogging for profiling (#1639)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1228, "deletions": 72}}, {"node": {"oid": "0c6e9f94d05f31556f4c2d6ae1abea8beb5014d8", "messageHeadline": "fix builds enabling onnxruntime_DEBUG_NODE_INPUTS_OUTPUTS (#2369)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 27, "deletions": 0}}, {"node": {"oid": "53ed36a3da0b23ad4346063eea6b14ff0c582afc", "messageHeadline": "Add helper to create output to minimize binary size. (#2365)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 33, "deletions": 29}}, {"node": {"oid": "aa37e2de8f9e2991341df437dfec40259b5e4e33", "messageHeadline": "Direct use python numpy array's memory if already contiguous.  (#2355)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 83, "deletions": 57}}, {"node": {"oid": "ed6da0d191196d5764190f39d0121d4fe923cf0a", "messageHeadline": "Implement cuda nonzero op. (#2056)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 257, "deletions": 0}}, {"node": {"oid": "3d3cf0e1598c35c018ba741ee4685612d562202e", "messageHeadline": "Openvino EP R3.1 onnxrt server (#2357)", "author": {"name": "avidiyal", "email": "41489828+avidiyal@users.noreply.github.com", "user": {"login": "avidiyal"}}, "additions": 87, "deletions": 2}}, {"node": {"oid": "599d72a94f6285254ba693c575c0396737c310ba", "messageHeadline": "Fix/test dim value of 0 handling in a couple of places (#2337)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 41, "deletions": 14}}, {"node": {"oid": "25b3c51661ceb1bc6bb5df0df8479391c0bd82f9", "messageHeadline": "Introduce PrimitiveType into a Type System along with an integer cons\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 2335, "deletions": 2318}}, {"node": {"oid": "fa30b1e75812eb56097a222e983ab170ab21caa4", "messageHeadline": "Set ElementType to String type of node metadata, instead of byte[] (#\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 6, "deletions": 8}}, {"node": {"oid": "7fcd75239346449e8bc2c1f3b7689be1c6337f0d", "messageHeadline": "Cuda Reverse Sequence Op, maping types of same size using same templa\u2026", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 230, "deletions": 0}}, {"node": {"oid": "080a0a3186cbfa10ae57b2876e4a51dfa706e6f4", "messageHeadline": "Nuget pipeline changes (#2305)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 223, "deletions": 497}}, {"node": {"oid": "5a3ea7469aa63539a23128339a8cf89f77c355a7", "messageHeadline": "Remove unused initializer from GraphProto as well as name_to_initial_\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 48, "deletions": 12}}, {"node": {"oid": "da3c0ba14b3925eb5b8abaf9b2e13f2321d0b29c", "messageHeadline": "implement CPU contrib OP Attention (#2333)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 313, "deletions": 59}}, {"node": {"oid": "b539cc74c71bf901aae6eb1515dfbdc11959d222", "messageHeadline": "Add FastGelu Cuda Op for Gelu and Add bias fusion (#2293)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 430, "deletions": 6}}, {"node": {"oid": "259bff8cf1ec21c584a5d987bab481f3bf1d0d37", "messageHeadline": "Layer Normalization Fusion  (#2319)", "author": {"name": "liuziyue", "email": "ziyue.liu7@gmail.com", "user": {"login": "liuziyue"}}, "additions": 354, "deletions": 0}}, {"node": {"oid": "553537ed52f23a93f145171fc41f4cf5c51e9635", "messageHeadline": "Add CUDA GatherElements kernel (#2310)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 357, "deletions": 27}}, {"node": {"oid": "6651d2f66259fc4723fc3855c9d7eb13b1af59e8", "messageHeadline": "Make elementwise op run 4 items per thread (#2335)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 167, "deletions": 123}}, {"node": {"oid": "ba0e7daf20500b758bb49d26290046fb9e656e7d", "messageHeadline": "update dockerfiles/README (#2336)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "0f1e24f4a9644c08781f5cbcb58cbb649edec5ee", "messageHeadline": "[NupharEP] tensorize int8 GEMM for avx (#2142)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 197, "deletions": 71}}, {"node": {"oid": "58e6aaa414d05293f68e8480f4bc0c4980e4fa63", "messageHeadline": "Fix crash in releasing TLS from CUDA EP dtor (#2329)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 29, "deletions": 17}}, {"node": {"oid": "c0b892686336c67b71ca35d4d0bbacb4b18d1f34", "messageHeadline": "implement CPU contrib OP EmbedLayerNormalization (#2332)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 244, "deletions": 37}}, {"node": {"oid": "06a6d74a67bb581a99c5c6aa828f34be832eb955", "messageHeadline": "update ngraph dockerfile. add python lib location to LD_LIBRARY_PATH \u2026", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 14, "deletions": 10}}, {"node": {"oid": "ace19129b93ecd856af77a4c9c66be870dc8c0e0", "messageHeadline": "MCR Docker Images v1.0.0 refresh (#2302)", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 9, "deletions": 8}}, {"node": {"oid": "151075790d59455ec654a46dcb65bd573a756878", "messageHeadline": "[OpenVINO-EP] Update to latest version: OpenVINO 2019 R3.1 (#2308)", "author": {"name": "Patrick Foley", "email": "psfoley@gmail.com", "user": {"login": "psfoley"}}, "additions": 223, "deletions": 386}}, {"node": {"oid": "db454beacfc8612b9da0d44d793bbc6b328fdd93", "messageHeadline": "TensorDesc::Placement test failure - cherry pick Vibranium fix. (#2328)", "author": {"name": "Dwayne Robinson", "email": "dwayner@microsoft.com", "user": {"login": "fdwr"}}, "additions": 31, "deletions": 8}}, {"node": {"oid": "67ec626d88679fa2847b2caba0c8c07945ecf9ad", "messageHeadline": "Copy blocks in Slice when possible (#2312)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 321, "deletions": 214}}, {"node": {"oid": "104f3b2a5968c6d66a396df841b5ff51bf31f4f2", "messageHeadline": "Exclude candy from CUDA tests", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "143ae98a377b99167d5a9417e9da998313f7d07a", "messageHeadline": "Fix a bug in onnxruntime_pybind_state.cc when TENSORRT is enabled (#2\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "8a102c6e99555c22be63a8c754a65c63b501d5c5", "messageHeadline": "apply eigen patch only for ACL.", "author": {"name": "George", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "5ce4d4fc494578ab3611c0e1136204d027b7d267", "messageHeadline": "Fix a test failure when it runs on FreeBSD", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "035913d42f91593d7f8457222a20e94cdd6f2e44", "messageHeadline": "Support int32_t for Reduction (#2317)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 160, "deletions": 30}}, {"node": {"oid": "d5c36bfff285eba4a0ef7c6b7dabb5b3bf6c5c2e", "messageHeadline": "Updated links in docs (#2303)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "556bae17a527a3389d08bce360f9671c2ebd71a0", "messageHeadline": "Fix versions table (#2309)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "cba93f7c8d3d71229149f9bcd9df26659d96f0ec", "messageHeadline": "fix Gelu CPU: remove MayInplace() declaration (#2306)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "204a6872d352ee4fa049e059ba9b86c56a8cdf6f", "messageHeadline": "remove unused param 'input_count' in ConcatImpl (#2304)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 0, "deletions": 9}}, {"node": {"oid": "a6b2c9fc09f601111e6d96b27ea13d3c2186ad58", "messageHeadline": "Fix mask in EmbedLayerNormalization (#2300)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 82, "deletions": 1}}, {"node": {"oid": "6e65dcf588ac0885d28c75a4a4c1895076e2e400", "messageHeadline": "[NupharEP] symbolic_shape_infer improvements (#2299)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 135, "deletions": 75}}, {"node": {"oid": "bc85d43809c7618f32a088162380e6b9f9e6cefc", "messageHeadline": "Dump cuda tensor data (#2243)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 20, "deletions": 13}}, {"node": {"oid": "7a5de9c958bd70a81794681cbe4370067faaa42b", "messageHeadline": "Add a python script with a number of helper actions for creating/edit\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 175, "deletions": 0}}, {"node": {"oid": "358b517d49af60890bb1901f2d621d0943ad70f5", "messageHeadline": "[v2] Add ACL (Arm Compute Library) execution provider (#2258)", "author": {"name": "mikecaraman", "email": "56535429+mikecaraman@users.noreply.github.com", "user": {"login": "mikecaraman"}}, "additions": 1479, "deletions": 15}}, {"node": {"oid": "bf7fa091cc2abca5eb7069b00436dbace12312d2", "messageHeadline": "NonMaxSuppression cuda implementation (#2082)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 669, "deletions": 28}}, {"node": {"oid": "67755adfd840f31209b46ffd6d0263ea129d88ef", "messageHeadline": "Bug Fix: NodeArg class has a move constructor but doesn't have a move\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "d6849bd26cf54bfb95f2e1301b291e2128f96f19", "messageHeadline": "Rashuai/cuda top k (#1919)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 244, "deletions": 0}}, {"node": {"oid": "4bcd8bfca1eaa791970caaa18c719bb31a8d0627", "messageHeadline": "Fix CUDA Reduce ops (#2268)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 149, "deletions": 3}}, {"node": {"oid": "a5da5ff6f47b12c8932373bf9fec9da43ecd6442", "messageHeadline": "Remove onnxruntime_USE_EIGEN_THREADPOOL cmake option", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 10}}, {"node": {"oid": "ff64d1f55b272bf4dfeff26b55890e0c6d432405", "messageHeadline": "Relax check for optimized model saving (#2291)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 8, "deletions": 15}}, {"node": {"oid": "ecfbb1bb99430f6f057734c92d072f1eb5fefdbc", "messageHeadline": "Add missing guards to profiling calls (#1374)", "author": {"name": "Maik Riechert", "email": "marieche@microsoft.com", "user": {"login": "letmaik"}}, "additions": 13, "deletions": 3}}, {"node": {"oid": "aa041026e362f0728232b4aeec3a23a48ed3d70f", "messageHeadline": "update Dockerfile.openvino (#2286)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 18, "deletions": 29}}, {"node": {"oid": "427e6278051fd8c687f3ef523a43f15b57e97d57", "messageHeadline": "Support for the Expand op with constant shape inputs (#2278)", "author": {"name": "Tomasz Do\u0142bniak", "email": "tomasz.dolbniak@intel.com", "user": null}, "additions": 4, "deletions": 0}}, {"node": {"oid": "e18c9582a814b0d0d8875becd9198590be238286", "messageHeadline": "[NupharEP] performance improvements (#2283)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 403, "deletions": 102}}, {"node": {"oid": "63e996163793188440ba0b15fac21daed9d36cc2", "messageHeadline": "fix typo", "author": {"name": "zhijxu", "email": "Zhijiang.Xu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "8dabe0502bb17d180f9f28a3b602dc64fac878ac", "messageHeadline": "merge two RUN to avoid making docker image too larger", "author": {"name": "zhijxu", "email": "Zhijiang.Xu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "7b11f05a97077a2286bdb44c42f1f2f431ddc745", "messageHeadline": "Update version number", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "2172a9e5ed549bb981da3f6cc16724d25dfeeb3d", "messageHeadline": "Fix an issue in the nuget run tests scripts", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "138a7f194ee0b50fdcfef40f9f2f547d502be3c2", "messageHeadline": "Add cleanup step", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 34, "deletions": 0}}, {"node": {"oid": "002d9c335b35dc304c0463bf5c6b04aa99939524", "messageHeadline": "Add node and op type info to error message if there's a type or shape\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 31, "deletions": 1}}, {"node": {"oid": "8d231a32f2e57ce6377a26279ee6f1a443e0fe6a", "messageHeadline": "Remove the libc version check in C# code (#2282)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 0, "deletions": 30}}, {"node": {"oid": "983a616bda3d2a8e4ea8a46e730ddae552a06638", "messageHeadline": "Revert to using opset 7 as the default for OpTester. Add explanation \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 86, "deletions": 64}}, {"node": {"oid": "47f40ca204a9a319d7368b81e312da40f562d427", "messageHeadline": "Cleanup add/remove of initializer (#2274)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 101, "deletions": 53}}, {"node": {"oid": "09eb8ff8b88ef96fd65575115ff04f388f77af49", "messageHeadline": "make sure samples run (#2276)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 17, "deletions": 10}}, {"node": {"oid": "d1096b524f03538df4437685c6cfac723e5ea887", "messageHeadline": "Fix fns candy build error on Linux", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "b0359b5622a893f5906df4b0f5cfc6ee30cc0844", "messageHeadline": "register where kernel (#2245)", "author": {"name": "Nathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 9, "deletions": 0}}, {"node": {"oid": "5611a528f6862ee389b77ba71bcc9e82b17e7682", "messageHeadline": "onnxruntime_perf_test: Add -u option to save optimized model (#2227)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 10, "deletions": 1}}, {"node": {"oid": "3ecdd985cb72323621478fbd40547946bbdeb3c4", "messageHeadline": "MNIST Sample Fix (#2259)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 23, "deletions": 10}}, {"node": {"oid": "0b88eff43af957351ce819a680db6ea95d5dcab4", "messageHeadline": "add docker file to build onnxruntime with different execution providers", "author": {"name": "zhijxu", "email": "Zhijiang.Xu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 162, "deletions": 0}}, {"node": {"oid": "3926ce33f4ab6d6a6772ca40c452e8f3e3a142b9", "messageHeadline": "add --parallel to speed up compiling source code", "author": {"name": "zhijxu", "email": "Zhijiang.Xu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "4eda08534fa2e8a42bf131c8428979355ef13021", "messageHeadline": "make smaller step in Dockerfile", "author": {"name": "zhijxu", "email": "Zhijiang.Xu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "ce23d628a5dee7d22760668b52f9a9cfc1179438", "messageHeadline": "fix bug in cmake/onnxruntime_server.cmake", "author": {"name": "zhijxu", "email": "Zhijiang.Xu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "be7c24247fd24961d86d1e3d640ebb41a5736f3e", "messageHeadline": "register execution provider when onnxruntime server creating sessions", "author": {"name": "zhijxu", "email": "Zhijiang.Xu@microsoft.com", "user": {"login": "zhijxu-MS"}}, "additions": 35, "deletions": 1}}, {"node": {"oid": "1fa956fb3f959b69ea7cf5f5b0fab2ace663fc25", "messageHeadline": "Undo integration test skip (#1917)", "author": {"name": "pulkittomar", "email": "pulkit.tomar@outlook.com", "user": {"login": "pulkittomar"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a2596b706be76265e25063a77aceacca770092f8", "messageHeadline": "FreeBSD compatibility patch.", "author": {"name": "Yuri", "email": "yuri@rawbw.com", "user": {"login": "yurivict"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "88c58c19d476dc3d8534e2518ce88977e41450bd", "messageHeadline": "Improve code readability and performance. (#2257)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 41, "deletions": 26}}, {"node": {"oid": "ce14b07b1c455b4444fc9924a501d2f23df17cfc", "messageHeadline": "Fix the GPU nuget pipeline failure (#2255)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c7599d0705a1f60ea2709e65647b67db220691b8", "messageHeadline": "subgraph parse error fix (#2254)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 20, "deletions": 19}}, {"node": {"oid": "6a0ee7eff6ce42935c73e52f88fb6d48eb1ea53c", "messageHeadline": "Fix model path marshalling in csharp, and re-enable the pretrained mo\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 16, "deletions": 10}}, {"node": {"oid": "8be48f47dd64a111f859c54616414a9dc126f66d", "messageHeadline": "Added Graph::ReplaceInitializedTensor() function. (#2230)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 119, "deletions": 2}}, {"node": {"oid": "7b4e6636f8c06abb65bb283bd8ededafa398d614", "messageHeadline": "Add default params to Env C++ constructor (#2246)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "63acd4e89b8368ca35785fd2a0ec17ab6218ea96", "messageHeadline": "Adjust the nGraph EP to the newest CI test data (#2180)", "author": {"name": "Tomasz Do\u0142bniak", "email": "tomasz.dolbniak@intel.com", "user": null}, "additions": 7, "deletions": 6}}, {"node": {"oid": "e250e51e652a1bf69e2bc7e423391899ccc682a6", "messageHeadline": "clear cache when moving to next channel (#2253)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 55, "deletions": 3}}, {"node": {"oid": "7494500221c1de4beee9658feb38a3d1d17737a5", "messageHeadline": "Fix csharp CXX sample (#2251)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "f7b4bc15e1d0120a68bc3f8909c6b002a9a5d1a2", "messageHeadline": "Updated documentation for VAD-F (#2248)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 77, "deletions": 73}}, {"node": {"oid": "20e6a2b6da536f865523d9aff0bd81466d8fabf0", "messageHeadline": "Disable optimizers for OpTester operator unit tests  (#2237)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 20, "deletions": 13}}, {"node": {"oid": "77d8d6f76756318a70826181d657c11b13067485", "messageHeadline": "Remove the OrtApiBase base_ member from OrtApi (#2242)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 0, "deletions": 5}}, {"node": {"oid": "6a27cb5ad6f280ed7e219a568712ba296006a5d8", "messageHeadline": "Fixed tensor reference to const data and cleaned up Env API. (#1979)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 641, "deletions": 176}}, {"node": {"oid": "557243dce9e25c19f3edbac540537d96a5a75a5f", "messageHeadline": "Add handling of zero dim to broadcasting and a few other ops (#2217)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 830, "deletions": 287}}, {"node": {"oid": "fef721c4f24d137a509786058296c2981686ad51", "messageHeadline": "Update build instructions. Make samples build and run. (#2244)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 63, "deletions": 27}}, {"node": {"oid": "bede664af7edeb5e463f2683a24d0d639f145d71", "messageHeadline": "mimalloc allocator (#2071)", "author": {"name": "kile0", "email": "kile@microsoft.com", "user": {"login": "kile0"}}, "additions": 274, "deletions": 123}}, {"node": {"oid": "90858b732a1d27657fc1424c91ccbfdf46082437", "messageHeadline": "handle Concat-11 (#2240)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 32, "deletions": 15}}, {"node": {"oid": "eb75388cd542403d1d95a28049e9c308ff39986a", "messageHeadline": "Minor optimization: don't log all node names if they've all been assi\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 9, "deletions": 5}}, {"node": {"oid": "41d55ea274493962ff12932c652a8dc7d0c1b60e", "messageHeadline": "Update the GraphProto for subgraphs when saving the Graph.  (#647)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 21, "deletions": 11}}, {"node": {"oid": "6fca8b0a94f54a63fdfb162a99059edacfe47565", "messageHeadline": "Move CXX API global into the header (#2228)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 225, "deletions": 275}}, {"node": {"oid": "90b6ff927f3e4aa94541d1bbfb5f73412f27f032", "messageHeadline": "Implement EyeLike Cuda Kernel to improve performance (#2139)", "author": {"name": "Nathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 199, "deletions": 0}}, {"node": {"oid": "ac9425c1195cb1fee3c672d34fd995837d776b11", "messageHeadline": "Update ORTSRV integration test model path (#2225)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 10, "deletions": 9}}, {"node": {"oid": "4b62241c777a7c16570ed072c250ba9710b76381", "messageHeadline": "Update ONNX to 1.6.1 (#2235)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "81d901cb600cfd91904c3d859afc8c7fcd0d5645", "messageHeadline": "remove nuphar scripts (#2233)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 0, "deletions": 42}}, {"node": {"oid": "ab39f7ec998b96fb824bf829984314586785f3e3", "messageHeadline": "Jeffbloo/fix dml rnn failures (#2234)", "author": {"name": "Jeff", "email": "38966965+jeffbloo@users.noreply.github.com", "user": {"login": "jeffbloo"}}, "additions": 11, "deletions": 9}}, {"node": {"oid": "041a1cab54493d3d32e6a6a8b4981add8c1419ed", "messageHeadline": "update libmkldnn.so version (#2220)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "6ac3e31047dbe9a661d4e75bde75352b1d725104", "messageHeadline": "Avoid printing a misleading message exposed by a previous commit + re\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 22, "deletions": 8}}, {"node": {"oid": "04b8a4bb34f4b6a7770723ea8b5917f1f1168620", "messageHeadline": "Log node placement info in verbose logging mode + turn on verbose mod\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 35, "deletions": 4}}, {"node": {"oid": "384c686f4015e3e92a555156aedbc35a8adebf21", "messageHeadline": "Update README and other files with the correct cuda version used for \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "ff2c4a4d7e8cb10cc5005d35812c5e6a2faa1499", "messageHeadline": "Added more input data types for pad (#2179)", "author": {"name": "Negin Raoof", "email": "neginmr@utexas.edu", "user": {"login": "neginraoof"}}, "additions": 295, "deletions": 153}}, {"node": {"oid": "b23cc04f347e99d57b7567cae663ad5585920e2c", "messageHeadline": "Handle the case when scales is constant but size is 0 (#2218)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 32, "deletions": 2}}, {"node": {"oid": "856c6cae0a98c690f99f94058b5e7640deb92429", "messageHeadline": "Edgchen1/endian utils (#2181)", "author": {"name": "edgchen1", "email": "18449977+edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 260, "deletions": 104}}, {"node": {"oid": "3507197599c5a2ceb053466e0ec92e867c8c8058", "messageHeadline": "The Conv+BN fusion is missing a check that the BN node does not have \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 58, "deletions": 0}}, {"node": {"oid": "b3446703516fed059a0272e6b685df6acbd39fcb", "messageHeadline": "Updated Perf Test readme (#2212)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 52, "deletions": 14}}, {"node": {"oid": "5eb42f4452aea2f3ce8277876526e75f7d797962", "messageHeadline": "Add uint8 support for BitShift operator (#2214)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 42, "deletions": 26}}, {"node": {"oid": "91122a2cf51516b89f483911d00a30da4315a2b0", "messageHeadline": "Fix GELU fusion (#2213)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 79, "deletions": 53}}, {"node": {"oid": "aef055ebe8e09bd1df9c1456922b062c50b52e41", "messageHeadline": "Update nuget pipeline to use CentOS6 (#2211)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 172, "deletions": 86}}, {"node": {"oid": "303a78c3012d12f9034d32ecce4f9680eff44314", "messageHeadline": "Update Python documentation (#2210)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 7, "deletions": 45}}, {"node": {"oid": "e8ec52814f80e5d49f9c7b5466e557f6f0552b56", "messageHeadline": "Remove some node tests from exclusion (#2168)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 17, "deletions": 55}}, {"node": {"oid": "cdcca43aa3903a961d2196764d25004e866aab97", "messageHeadline": "Skip GemmNoTrans_f16 test for CUDA if the hardware does not support fp16", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "d1970794736f65c2b848bc05d8bfd72fed1c4fc8", "messageHeadline": "quantization script updates (#2208)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 430, "deletions": 241}}, {"node": {"oid": "b6b44c90ac6fc7eca823e0c926c30a86214bd85c", "messageHeadline": "Fix MaxUnpool behaviour when output_shape is provided and doesn't mat\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 85, "deletions": 182}}, {"node": {"oid": "6699c190103a1b387bd9a172a847c7f27c45270d", "messageHeadline": "Add script to find if optimizer updates are required due to onnx oper\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 134, "deletions": 0}}, {"node": {"oid": "dc5efbf5cef9996b43e3f440b42b119d0a3897c8", "messageHeadline": "enable exclude outside for resize op (#2203)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 66, "deletions": 11}}, {"node": {"oid": "3cda9f717bb780e2af9a0057c8d90db06869927c", "messageHeadline": "Relax shape inferencing error handling if model uses an old opset (#2\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 212, "deletions": 88}}, {"node": {"oid": "02dc3a9dcb9889efbd0e78c53b1932cc03e33b1e", "messageHeadline": "build break for arm64, adding advapi32.lib (#2206)", "author": {"name": "Paul McDaniel", "email": "paulm@microsoft.com", "user": {"login": "walrusmcd"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "fcf50ca0810b3d68a8db96ace8d41382547173c5", "messageHeadline": "Fix nuget mklml pipeline (#2204)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "5c86889beba9ed7481e9db1095c35a39b778558d", "messageHeadline": "Fix linux build issue with debug dump of shapes and data. (#2202)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 23, "deletions": 18}}, {"node": {"oid": "07e9f500da0d01b984f3d4e41770957bca06b080", "messageHeadline": "Add back Keras Mask RCNN to CUDA test exclusion (#2201)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "aae18a3fe36d0b8e85272382822cdf76a0e0ac6a", "messageHeadline": "Upgrade onehot to OpSet 11 (#2185)", "author": {"name": "Nathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 106, "deletions": 29}}, {"node": {"oid": "69970d1f2a77c49a9b8cac91cadf6678edc0189d", "messageHeadline": "Include the new Privacy.md file in all release packages. (#2200)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "cf7ee5f6e8bcf9d039e75ed832cba04b9794cff9", "messageHeadline": "Support negative axis in unsqueeze elimination (#2158)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 34, "deletions": 12}}, {"node": {"oid": "33c639a0224011736c1d6135953d9fceb84086ca", "messageHeadline": "Slice elimination support for opsets 10 and 11 (#2171)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 685, "deletions": 30}}, {"node": {"oid": "cff7879d898220f5ea86d317209ac21bd4753983", "messageHeadline": "Update C API pipeline to use CentOS 6 (#2198)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 56, "deletions": 55}}, {"node": {"oid": "0dd781fd57d4b34a39490c63b6d85cb9376a3549", "messageHeadline": "Perf tuning doc update with latest API (#2128)", "author": {"name": "Nathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 26, "deletions": 16}}, {"node": {"oid": "a9f01a5f29d96b3e924afff6966aa4e45c6372ed", "messageHeadline": "Fixed node index remapping issue in TensorRT graph partitioning (#2155)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 120, "deletions": 2}}, {"node": {"oid": "7b18bd563f74bc9198cb79d12535927384a30e2b", "messageHeadline": "Commit formatting and end of lines (#2161)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 411, "deletions": 411}}, {"node": {"oid": "763af5d42a6c069c0756e80007a5718dcca54271", "messageHeadline": "Remove Keras Mask RCNN from CUDA test exclusion list (#2196)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "18b192a45bf7ca3752cecd074786e58d8f4d3719", "messageHeadline": "Implement Range Cuda Kernel to improve performance  (#2148)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 189, "deletions": 0}}, {"node": {"oid": "7efc9bdcc775c6ae1dc75d5216fc2c81b8a91626", "messageHeadline": "Some condition fixes on nuget pipeline, to get it green (#2195)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "836d22cd4c00a153dfce20d7afc8d1ba757fd318", "messageHeadline": "Update readme.rst for pypi, change documentation style (#1663)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 6, "deletions": 93}}, {"node": {"oid": "ac3d2ad8970b19f6f620ec9ccbe303a8e2adb65f", "messageHeadline": "Implement ConcatFromSequence (#2106)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 369, "deletions": 71}}, {"node": {"oid": "d1159b70084bc7543ff7a658bd4de1e085f4222e", "messageHeadline": "Adding platform telemetry (#2109)", "author": {"name": "Paul McDaniel", "email": "paul_mcdaniel@hotmail.com", "user": {"login": "walrusmcd"}}, "additions": 551, "deletions": 35}}, {"node": {"oid": "b1096424f03a1016bb06d0d59719798f554c373e", "messageHeadline": "mkldnn cleanup. add EP_FAIL paths for unsupported mkldnn nodes. (#2182)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 113, "deletions": 157}}, {"node": {"oid": "acec4b446f2cd089f19ba8c6a46d88a7d5ec99f5", "messageHeadline": "Make CentOS 6 CUDA build and run (#2159)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 161, "deletions": 86}}, {"node": {"oid": "96b33f45976724eb0a92fa4eb39018f2f4f74315", "messageHeadline": "Add support for double to SplitToSequence op to allow test_sequence_m\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "f8c30b8aa9e04d00f60d5839541c9f2b523ab1a8", "messageHeadline": "Disable DML builds for now until further investigation since the test\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6d4682581e234a88d7c06ed2c337b641db0486e0", "messageHeadline": "resize opset11 (#2088)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 992, "deletions": 159}}, {"node": {"oid": "021073b5e52699a4413111ed38a29accfa571f91", "messageHeadline": " Update python packaging pipelines (#2167)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 24, "deletions": 67}}, {"node": {"oid": "d15021f42b704e054f2c909f142ea2596c3bf852", "messageHeadline": "Disable some tests for DML and Android EP pending investigation. (#2187)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "35dae992f17a1a96c21721da8bc11ee482195e2b", "messageHeadline": "Fix nuget gpu ci test error (#2164)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "ecf5ae8b76ec275ba2b3b61d740c46eb56b7d1fd", "messageHeadline": "Askhade/disable csharptests (#2172)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "62281ff3b9229d27fc8448666148925db8d7dd1f", "messageHeadline": "Add empty tensor support to the DML EP (#2163)", "author": {"name": "Patrice Vignola", "email": "vignola.patrice@gmail.com", "user": {"login": "PatriceVignola"}}, "additions": 231, "deletions": 64}}, {"node": {"oid": "5eb4e81f8025be757cc5eef3e8d6b3de30d1e8e7", "messageHeadline": "move some optimizers to level1 (#1566)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 48, "deletions": 55}}, {"node": {"oid": "47bc35126513a59ee0378ed19506930c77dc506c", "messageHeadline": "Fix wrong usage of GetScratchBuffer, like: (#2101)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 9, "deletions": 4}}, {"node": {"oid": "fc3c16840205506162448203cdd5cd99804042ca", "messageHeadline": "Graph Optimizations Doc (#2050)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 136, "deletions": 3}}, {"node": {"oid": "13f8b49d58bd58c3e12c42779bd2dcbd0d8760e7", "messageHeadline": "Fix kernel registry bug (#2137)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 176, "deletions": 86}}, {"node": {"oid": "2bf1778a5c29a7033bed9cca9660fe76812bd770", "messageHeadline": "Fix ORTSRV docker build (#2160)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "00e2d1c60457a44dcc85f6c428e9dd660ea5b1f5", "messageHeadline": "update (#2140)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 36, "deletions": 2}}, {"node": {"oid": "0e6ac2961ece44fcd447d6901157b2d1e0eae90f", "messageHeadline": "Adding a line beak to BUILD.md (#2156)", "author": {"name": "Clarke Rahrig", "email": "clarker@microsoft.com", "user": {"login": "crahrig"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "f6c82358e937e9778022c3b08fa4ebce552adb27", "messageHeadline": "Use built-in wstring_convert on Android (#2095)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "86af54ded8c812d847250d3d1806da315a81c39e", "messageHeadline": "Add roadmap file (#2127)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 147, "deletions": 0}}, {"node": {"oid": "72110d35083cfde12a335d37be0e54218d3cd3f3", "messageHeadline": "Patch for the MKLDNN v1 segfaults (#2145)", "author": {"name": "Tomasz Do\u0142bniak", "email": "tomasz.dolbniak@intel.com", "user": null}, "additions": 33, "deletions": 4}}, {"node": {"oid": "4006e2960173af5ea903ae3573a77558ef9e8fbd", "messageHeadline": "Implement SplitToSequence op. (#2131)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 402, "deletions": 0}}, {"node": {"oid": "70e7eaf1e863245caf06f9916fec15a81a85e44c", "messageHeadline": "Update DML transformers with the new Graph API and re-enable DML in t\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "95fef56dc80806d4409f308639ff17e0fc44fc73", "messageHeadline": "MKL-DNN EP (#2149)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 246, "deletions": 215}}, {"node": {"oid": "6445e7182c505b6ce6c9c6cc81a4fc67bc802b10", "messageHeadline": "Disable DML build temporarily until we fix the removal of the IsNodeO\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "b0f8ec7a7d4438f4b6eadf2f8f41d49b4e2212f4", "messageHeadline": "Script for converting BERT model for performance optimization (#2037)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 947, "deletions": 0}}, {"node": {"oid": "ec136ac60f9052069d82bd8393ba0d44407c15aa", "messageHeadline": "Documentation Refresh (#1990)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 513, "deletions": 458}}, {"node": {"oid": "3fcb4ee7d4e59ff095359110fbf467fcd9ee0f04", "messageHeadline": "Refine optimizers (#1407)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 640, "deletions": 426}}, {"node": {"oid": "485c24b62da3a280e5fb2cbe732729427a9daf29", "messageHeadline": "MKL-DNN 1.0 (#2134)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 1145, "deletions": 2442}}, {"node": {"oid": "6857bb8abaf9c214447e8dbef139cf67205d8e27", "messageHeadline": "Fix bug in GatherElements (#2130)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 109, "deletions": 51}}, {"node": {"oid": "7ef02f14d21b1cce77dd27e8e51fbb7389113273", "messageHeadline": "Add missing test model file for symbolic dimensions (#2123)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "4090d0d0de040794f0ca181f780600f3b2b9834b", "messageHeadline": "Add DirectML Execution Provider (#2057)", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 29239, "deletions": 36}}, {"node": {"oid": "b101f1bcee047430616a25ba0c227bd755e84dc4", "messageHeadline": "Nuphar: Fix a bug in weight layout where read may go out of bound (#2\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5c2803f2d537b3cc9e736651436d400d60409c59", "messageHeadline": "various fixes for shape inference script (#2124)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 80, "deletions": 43}}, {"node": {"oid": "95ab5ad39f4762f9bc6ab92c051253d725f0656f", "messageHeadline": "Support non-spatial mode in BatchNormalization (#2092)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 181, "deletions": 48}}, {"node": {"oid": "2536553136cc3fb09653fb06532013b7e3abb5e6", "messageHeadline": "use cublasHgemm for Volta GPU (#2074)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 68, "deletions": 66}}, {"node": {"oid": "8c5db7f9734fa0ce49f0ddf6ebead3653bb6b28f", "messageHeadline": "use legacy stream mode (#2076)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 4}}, {"node": {"oid": "80d09f0c59e551c8e7d9c13b14127268e0638a3d", "messageHeadline": "Allow creation of empty tensors in c# (#1976)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 61, "deletions": 3}}, {"node": {"oid": "640f71c91b455e2a40ab95b8fef562f6ab4cf6d8", "messageHeadline": "Enable Gpu multi-device test for CUDA EP and Trt EP", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 54, "deletions": 12}}, {"node": {"oid": "f93be8af90fa68cad0e02395ae4b94e3791b0ed0", "messageHeadline": "Update nGraph to version 0.26 (#1965)", "author": {"name": "Tomasz Socha", "email": "tomasz.socha@intel.com", "user": {"login": "tsocha"}}, "additions": 109, "deletions": 302}}, {"node": {"oid": "91db840b6b9336f691d6c73c58ef5a8f9688c302", "messageHeadline": "Introduce execution mode enum for clarity and extensibility; Change P\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 253, "deletions": 225}}, {"node": {"oid": "5558b8077483c5d020ea2e05373fa5394cb355cf", "messageHeadline": "clean up ubuntu docker scripts (#2103)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 28, "deletions": 35}}, {"node": {"oid": "9363c14d23e444e629167825a1f5a44b55aa9eaf", "messageHeadline": "Fix a bug in nuphar settings where existing options cannot be overrid\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 68, "deletions": 67}}, {"node": {"oid": "b829d5532047f2424930fc2337cd2d954f9f987d", "messageHeadline": "Fix invalid logic that ran past end of nodes and double increment. (#\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 11, "deletions": 10}}, {"node": {"oid": "eb24617d2e195f3e1d4c419065b0bdfca6baa354", "messageHeadline": "Add ability to get symbolic dimension info for graph inputs and outpu\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 258, "deletions": 80}}, {"node": {"oid": "20515363e5d12cde99f06e441f993905850462e6", "messageHeadline": "Add int32 and int64 types for Equal(11) (#2112)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "50faab308b7df1ba293e17e4373c78e6c325db55", "messageHeadline": "Remove 'Ort' prefix from OrtAddFreeDimensionOverride for consistency.\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "e8e33977dad1a324411749fce6f0c517948df873", "messageHeadline": "Ryanunderhill/customop dll (#2002)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 100, "deletions": 36}}, {"node": {"oid": "c24d7a8a0a95002cfde70db8ffdac1a1d95b3ca7", "messageHeadline": "Update eigen to the latest version (#1910)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "bdfff800ea47259d6d72ad12b128a9af5447d49a", "messageHeadline": "Move access to intra-op threadpool into OpKernelContext. (#2091)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 161, "deletions": 168}}, {"node": {"oid": "368bdfd936a1522ed9b21fe3471d01d074d8847d", "messageHeadline": "Update README.md (#2070)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3b335c933f0f39d2641200171cd9e04525002914", "messageHeadline": "Fix issue that TRT not work for device other than device id 0", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "ffb94fd1704ea012565d3257ca8e224f2ea28205", "messageHeadline": "Fix bug with delayed allocation of If and Scan outputs. (#2024)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 107, "deletions": 49}}, {"node": {"oid": "ca1b88c069a152a867e82aa518c3a417139d8c8e", "messageHeadline": "Added support to infer Pad11 (#2085)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "8803f6fff445934afdded590f81fddae2bfe0868", "messageHeadline": "C# end to end test fix, and make end to end tests mandatory (#2079)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 81, "deletions": 7}}, {"node": {"oid": "a314402097877cd1ddb4aa3d692b9c6337f9f0aa", "messageHeadline": "Downgrade python gpu package to CUDA 10.0 (#2086)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 59, "deletions": 116}}, {"node": {"oid": "af9dbb70f2eb9ef0baffab2e3325087d57f8f5c7", "messageHeadline": "Introduce a separate check and conditional for AVX512BW build (#2083)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 43, "deletions": 28}}, {"node": {"oid": "2ba705ed99046c1935abdd838ff0855a05c9ee19", "messageHeadline": "Handle nodes with subgraphs in ORT function handling implementation (\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 166, "deletions": 73}}, {"node": {"oid": "2d4d0abd366a000ac404231aa6ce5317af004528", "messageHeadline": "Add support for output seq(tensor) in python binding and test framewo\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 771, "deletions": 135}}, {"node": {"oid": "ddbc2086e4ecf501216a81097d717ad58eb7f42e", "messageHeadline": "Add support for opset 11 Clip in optimizers. (#2059)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 371, "deletions": 28}}, {"node": {"oid": "a41c71cbf2f203728264320c75b1922438f63f57", "messageHeadline": "check and fix CUDA kernel launch errors in several OPs (#2047)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 95, "deletions": 77}}, {"node": {"oid": "b4a98aab789eec41cc6815dd24e8fa33c86f9a84", "messageHeadline": "change MatMulInteger/MatMulInteger16 fallback option (#2064)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "d186c19c454279ae0536e6361bcc5e649a5a1c17", "messageHeadline": "Add opset-11 TopK CPU kernel (#1912)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 686, "deletions": 269}}, {"node": {"oid": "8fda6593fe0b2d9759fe833aa9dd3b78b4ee84ee", "messageHeadline": "Update failing tests (#2038)", "author": {"name": "Colin Versteeg", "email": "cverstee@gmail.com", "user": {"login": "csteegz"}}, "additions": 31, "deletions": 11}}, {"node": {"oid": "57e00994256ec6394e635f30b085e17bae8d0594", "messageHeadline": "MLAS: Implement U8S8 GEMV kernels (#2069)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1679, "deletions": 28}}, {"node": {"oid": "eee9c5503004199a5a3e1f693bf7ad8d119e7204", "messageHeadline": "C++11 fix for memcpy_transformer_test.cc (#2061)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "cefae93305a45526dc90918270d43598d99c1761", "messageHeadline": "Add a test case for linearregressor (#1962)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 56, "deletions": 37}}, {"node": {"oid": "ccaf692ff2266c32b06b35031339539f4384f57e", "messageHeadline": "Run auditwheel for manylinux1 (#2063)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "cae571c7136328a2f86ab6ecbeb1d82a01e673c3", "messageHeadline": "Add a test for AVX512 compilation before compiling 512 asm (#2055)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 68, "deletions": 31}}, {"node": {"oid": "af8fe0f98004635e105fa5031f1c3de8381e86b4", "messageHeadline": "Replace make_unique in cuda_utils.cu (#2052)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "db0dd09deda1e611ce1ab5728f273336e2f42649", "messageHeadline": "Cleanup some aspects of the Initializer class used by optimizers (#2005)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 145, "deletions": 129}}, {"node": {"oid": "a00ca56ae10659d0a78c236d0dc8ab4f62baaf54", "messageHeadline": "Remove gcc from manylinux1 docker image (#2048)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "b82de794d5d0421cdd53b822603cdd6276cf4414", "messageHeadline": "Weba/update nuphar doc (#2026)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 16, "deletions": 0}}, {"node": {"oid": "f501b6e2349b69bfb2f8738fe009467b1abe5437", "messageHeadline": "pack pyop in nightly build (#2018)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 20, "deletions": 3}}, {"node": {"oid": "e9bed8b23b518b88bfd0480611a3ac295228c03a", "messageHeadline": "Change python packaging pipeline to use manylinux1 (#2035)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 139, "deletions": 56}}, {"node": {"oid": "3053af812c45e35092780ddbfa404f96a6b6ceed", "messageHeadline": "Fix a crash in deep_cpu_gru_op_test.cc (#2028)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "71b389322e3a91b81d02d76e39d60d6b25d5d39d", "messageHeadline": "Implement cuda scatter op. (#1991)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 332, "deletions": 1}}, {"node": {"oid": "a94c9bd88dc05a6d18c152722d4ee7eb80b6a409", "messageHeadline": "throw exception using dmlc::LogMessageFatal (#2033)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 7, "deletions": 5}}, {"node": {"oid": "19b0d0af87b2dfaca56462bf810c55c7e67bd9ab", "messageHeadline": "Enabled bool input type for Equal for op_ver 11 (#2034)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 13, "deletions": 0}}, {"node": {"oid": "203c2f5b5917e47c24f316a50e808547622e26a9", "messageHeadline": "updated reduce_ops for op_ver 11 (#2039)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "f13b66768a3345318e0681818dfe96f720dc39cb", "messageHeadline": "Fix build for gcc 4.8.5. (#2036)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "b70fc34faea396a8db099a17ed4c7dea22f7e71f", "messageHeadline": "Fix C# end to end tests in NuGet pipeline, failing for missing test d\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "b0feaef9de0eeab34480bec00f83b90fb3e0b5e1", "messageHeadline": "Update the C# pretrained model test to include opset9 and 10 models (\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 324, "deletions": 74}}, {"node": {"oid": "0bd807f3b3c4d06a83f536eae78411bbccb8aaec", "messageHeadline": "trt provider status return cleanup (#2032)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 17, "deletions": 19}}, {"node": {"oid": "b2c193752376e6c3a802f8165093d8602a4ce4ba", "messageHeadline": "Add EmbedLayerNormalization and SkipLayerNormalization ops for bert o\u2026", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 1341, "deletions": 2}}, {"node": {"oid": "8f7657fa328ad875ac66fa2496c2a4602bdcc8b5", "messageHeadline": "Ignore some gcc warnings (#1996)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 32, "deletions": 3}}, {"node": {"oid": "ea60469af551f94b9b0452d0687f5be2c8b8a5ff", "messageHeadline": "Support seq(tensor), implement 2 sequence ops that use the new type. \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 776, "deletions": 187}}, {"node": {"oid": "00e24ae4fe4dde51d0ab1adfd3070b18222c2eaa", "messageHeadline": "refactor Cuda Ops Sum, Max, Min, remove dup code (#1946)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 83, "deletions": 143}}, {"node": {"oid": "7b39f5090ce00157dbf2b84b78858b96a4864bd7", "messageHeadline": "Add Attention op for multi-head self attention in BERT (#1984)", "author": {"name": "Tianlei Wu", "email": "tlwu@microsoft.com", "user": {"login": "tianleiwu"}}, "additions": 906, "deletions": 1}}, {"node": {"oid": "7d2f0c79bd5bdbbfb368efdf90755e9fa4b540e6", "messageHeadline": "Bumped up to op_ver 11 for a bunch of Nuphar Ops (#2025)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 93, "deletions": 26}}, {"node": {"oid": "3c26ae5b6d5ab6024dffae0794ae0774e7e1109b", "messageHeadline": "ThreadPool fix for roialign and CropAndResize (#2020)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 19, "deletions": 2}}, {"node": {"oid": "4cdb95e4367fb5d3aba0ddc57c9c1be0962ad9bf", "messageHeadline": "Resort to sequential execution if the inter op thread pool ptr is nul\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 59, "deletions": 14}}, {"node": {"oid": "544e53e24eff9015f7ae2f302f693c41afadc0a0", "messageHeadline": "Update TensorRT to version 6.0.1.5 (#1966)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 320, "deletions": 141}}, {"node": {"oid": "4bb6385dcab5e6bb0af2bf494538b92fbdde1683", "messageHeadline": "Weba/merge ngemm (#2021)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 1181, "deletions": 190}}, {"node": {"oid": "0b5aac0a2e95e517cafb3f0507c264d2bc01b732", "messageHeadline": "fix python setup (#2022)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e8285a7996dcca617ae925ea8c18aaa0c59bd242", "messageHeadline": "Added GatherElements to Nuphar (#2016)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 127, "deletions": 2}}, {"node": {"oid": "1ba76c5f743e3209e88fa95199df1713b9bf33b8", "messageHeadline": "add support for empty version and score route (#1995)", "author": {"name": "Colin Versteeg", "email": "cverstee@gmail.com", "user": {"login": "csteegz"}}, "additions": 50, "deletions": 11}}, {"node": {"oid": "a9e04a29b321cad98f1972d2d1b9a0659dc22f63", "messageHeadline": "Ignore a test: ParallelExecutor.StatusPropagation (#2019)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2a2e6e6641472beaba2346113a76d06c3a19f6bd", "messageHeadline": "Handle nullptr for NodeArg.Shape() (#2009)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "f528da35f26be356c4e97cf94631ee91abbde67c", "messageHeadline": "Update ONNX to a newer commit (#2015)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "f5a8a23951cc5474e4a81b2f852c7916a44e8f90", "messageHeadline": "Replace std::regex with re2 bc CentOS std::regex is broken (#2017)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 38, "deletions": 12}}, {"node": {"oid": "e071a1249b48e5dbbedb1df580d66aaeda71f79a", "messageHeadline": "Android CI (#1600)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 16, "deletions": 9}}, {"node": {"oid": "bfa1b0e96e0196a7accc1ef168c9338fdeb335a3", "messageHeadline": "Fix logger regression (#2011)", "author": {"name": "Colin Versteeg", "email": "cverstee@gmail.com", "user": {"login": "csteegz"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "b322e072b982d2df5f0690da4e3b5319ba9d6d87", "messageHeadline": "added the overridableinitializers api (#1977)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 145, "deletions": 4}}, {"node": {"oid": "19873c70dc179c34a131ba9ee7fa9e886b5a4d34", "messageHeadline": "Implement Cuda Kernel of  Where Op (#1997)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 358, "deletions": 0}}, {"node": {"oid": "a6bf1d0ad8cfd6457b7b4cec74b84c5a24046fcd", "messageHeadline": "use mlaserf (#1999)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 6, "deletions": 2}}, {"node": {"oid": "fdbe365c37745556629ffe28e0c56e0dbce6646f", "messageHeadline": "Add BitShift operator (#1981)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 197, "deletions": 40}}, {"node": {"oid": "d5d1719c1f0beb4a6cc45fa588290c6b85a33540", "messageHeadline": "Fix integration_tests/test_main.py to have correct exit code (#2010)", "author": {"name": "Colin Versteeg", "email": "cverstee@gmail.com", "user": {"login": "csteegz"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "ace0b2ca1c64aa95d49781202950992a875ffa6c", "messageHeadline": "CentOS CI (#1998)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 32, "deletions": 6}}, {"node": {"oid": "534660bf2c00959794a6e184d8b25574bc40b19b", "messageHeadline": "Support opset-11 Range CPU kernel (#1980)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 198, "deletions": 229}}, {"node": {"oid": "a7414287a9ef2053323c755ea33de798ad60870c", "messageHeadline": "fix Conv/Add fusion with no bias (#1978)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 40, "deletions": 5}}, {"node": {"oid": "d80f324fd343baeb8f05e71f28199159d249bbdc", "messageHeadline": "Add Pad test to NGraph specific exclusion list (#2006)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "e58827fa62eb59c41363d9a13091863095d771b6", "messageHeadline": "Add Unique operator. (#1900)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 789, "deletions": 12}}, {"node": {"oid": "74517bb7425b52fd15fef5cf4d4f476c1987e764", "messageHeadline": "Support opset-11 GatherND CPU kernel (#1969)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 884, "deletions": 625}}, {"node": {"oid": "627f853a446205c880eccd06dbb1c3b77ec498c4", "messageHeadline": "Downgrade compiler to CentOS 4.8.5 (#1985)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 326, "deletions": 176}}, {"node": {"oid": "931975e3fe6bb6e4727c78af6854cb6090a54fa5", "messageHeadline": "Add opset-11 Pad CPU kernel (#1963)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 268, "deletions": 399}}, {"node": {"oid": "15138908e7ff72b542f681ad46e6bddc242e8a49", "messageHeadline": "Yanchen/nuphar/scatter elems (#1992)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 327, "deletions": 0}}, {"node": {"oid": "c86d17754a0d9dfa915c4dbce895aa642a323532", "messageHeadline": "Dockerfile for CentOS CI build (#1986)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 79, "deletions": 48}}, {"node": {"oid": "a1f8fe14637d68810df031a6b7ca54a7219091b4", "messageHeadline": "Add layernorm operator (#1967)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1082, "deletions": 38}}, {"node": {"oid": "6b6d0aa888a7487c3f64c22e5212f54672072973", "messageHeadline": "Fix the NGraph backend test exclusions (#1988)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 102, "deletions": 98}}, {"node": {"oid": "9959e84906259585cfc4cfbd3deeccdb60a2bf3d", "messageHeadline": "Gelu fusion - kernel and transformer (#1746)", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 377, "deletions": 64}}, {"node": {"oid": "b0665262c0531e3e12f0ce6558037b04644c34d7", "messageHeadline": "Fix how failing tests are defined. Merge of old PR didn't detect conf\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 10, "deletions": 8}}, {"node": {"oid": "c79195fb3b82083dbb6b5e3f3ed72521bc007659", "messageHeadline": "Add GatherElements CPU kernel (#1735)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 508, "deletions": 4}}, {"node": {"oid": "9f633c5bd9ecccdb0d02e0088d516d67a9dab7cb", "messageHeadline": "Update Cast op to use precision of 8 when casting floating point numb\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 56, "deletions": 32}}, {"node": {"oid": "9e975f64c33a7f86c40fea179fbc38196a1f34cc", "messageHeadline": "Add scatter_nd cpu (#1603)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 365, "deletions": 1}}, {"node": {"oid": "7c77a01ce7c9f3015cc7473d34856fc2dc380887", "messageHeadline": "OpenVINO dockerfile for ORT (#1892)", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 13, "deletions": 13}}, {"node": {"oid": "103b92889e5fdd286f742db6127b6233176fdb6a", "messageHeadline": "Opset-11 support (negative axis) for reduce ops (#1929)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 6401, "deletions": 5495}}, {"node": {"oid": "f9bf546e3c19d1739590a24cfdf3b08a503f8c73", "messageHeadline": "python session.run() fallback to CPU/CUDA provider for EP failures. (\u2026", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 82, "deletions": 40}}, {"node": {"oid": "622ea4248df1531eb527a12818743395474c47c1", "messageHeadline": "fix build instruction (#1970)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 39, "deletions": 7}}, {"node": {"oid": "a5e134405d9f8cfa3dc9ec012f50bafb9b95c806", "messageHeadline": "Support opset-11 Gemm kernels (#1923)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 92, "deletions": 37}}, {"node": {"oid": "31aff686e0f73d12fbc6b5313e439597f9798f75", "messageHeadline": "Register opset-11 CPU kernel for 'If' op (#1948)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 69, "deletions": 20}}, {"node": {"oid": "d1b1cdc5c49c2ec18d4cdcfbb1cf58b7ea18e556", "messageHeadline": "Replace GSL with GSL-LITE submodule and fix up refs (#1920)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 4245, "deletions": 617}}, {"node": {"oid": "983ed43f0a37056cb1866d3f9090df03a3199686", "messageHeadline": "Fix declarations for opset 9 versions of Gemm, MatMul and BatchNormal\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 75, "deletions": 28}}, {"node": {"oid": "b355193841e56433b117b33b73fa1bf063168a5a", "messageHeadline": "Add Date-time stamp in NuGet package versioning for appropriate order\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 78, "deletions": 68}}, {"node": {"oid": "aef49d1f2200a44ba2290c453c052ea9090b5679", "messageHeadline": "Update ops that had strides/dilations documentation updates to defaul\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 311, "deletions": 137}}, {"node": {"oid": "df472cbfbdf0ff47f1051639fe0216178638e6ab", "messageHeadline": "quantise: Don't error when initializer graph input is missing (#1872)", "author": {"name": "Jonny Shipton", "email": "tmvector@gmail.com", "user": {"login": "TMVector"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "e361174f7874f24e1c0492984b15cf3fe886fc4f", "messageHeadline": "Add nuphar python scripts to wheel, and notebook tutorial (#1952)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 961, "deletions": 132}}, {"node": {"oid": "4c995d3251b3c1826279ccfdb1701cfea74de721", "messageHeadline": "MLAS: add DGEMM support (#1953)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 6330, "deletions": 3992}}, {"node": {"oid": "9fc5598b7ed36ebdfd4bc05b7d073ae0da44efb2", "messageHeadline": "update nuphar ci llvm version and uncomment unit tests (#1954)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 3, "deletions": 13}}, {"node": {"oid": "611dd3ea0c41442f7ed5994fe5345ced2a546fd1", "messageHeadline": "update ort-tvm version (#1945)", "author": {"name": "baowenlei", "email": "40643946+baowenlei@users.noreply.github.com", "user": {"login": "baowenlei"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e4b8494eaa3ac3db1b677b6d8f13c4cc1c482ce7", "messageHeadline": "Add a test case for ml_commmon.h (#1947)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 87, "deletions": 0}}, {"node": {"oid": "02c122d6e478d69ee97b3ecdd0aac5cd4f76ba8c", "messageHeadline": "  Add OLive in perf tuning section (#1772)", "author": {"name": "Emma Ning", "email": "43255631+EmmaNingMS@users.noreply.github.com", "user": {"login": "EmmaNingMS"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "8df3e87b7063fc071c0d5fe35384d966915a6e87", "messageHeadline": "OpSet 11 Update for Neg Axis (9 Ops) (#1893)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 327, "deletions": 97}}, {"node": {"oid": "d370aad80c3fb995bd33e760b75f1edb10b3bcaf", "messageHeadline": "Register CPU kernel for NMS (#1941)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 33, "deletions": 3}}, {"node": {"oid": "174c142effc2cd73bc182f0847a688bce46accde", "messageHeadline": "Clean up code (#1939)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 6}}, {"node": {"oid": "1649374a5c551d3ba873fb309b76dfdce9e3c10c", "messageHeadline": "Some bug fixes and support for Gather/ScatterElements (#1940)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 30, "deletions": 12}}, {"node": {"oid": "ceaaff0f812dbef8b2c509bf82ef805af4810e20", "messageHeadline": "[OpenVINO-EP] Enabling VAD-F in OpenVINO Execution Provider (#1885)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 49, "deletions": 27}}, {"node": {"oid": "650fb8754b03ae89f6b0d456370ecb2dc901f4bd", "messageHeadline": "use MLAS for nuphar's pool ops (#1937)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 417, "deletions": 125}}, {"node": {"oid": "c3ffd1f47d3646addbdf32bdfdb8c790531f098a", "messageHeadline": "added continue on error for the linux cleanup step, to mitigate the b\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "2ecac41614349e2d3fbf123be134fde744c5b77a", "messageHeadline": "update python examples (#1935)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 73, "deletions": 37}}, {"node": {"oid": "a5a57a49c42d97434923d2603dfe50e8ed5fe723", "messageHeadline": "Changed ConvBase into a class member variable (#1927)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 442, "deletions": 403}}, {"node": {"oid": "7e22ed41b9317c66bc77ec4ad1fb8dbeb2b65f23", "messageHeadline": "Fix sample tests (#1926)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 39, "deletions": 36}}, {"node": {"oid": "09cdbe9d76636e8eae06b08fd8580e98f913c04c", "messageHeadline": "Update test data (#1932)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "67be535030bdf4a35c28078ea4549e3ca35e4693", "messageHeadline": "Fix a typo (#1934)", "author": {"name": "Mohammadreza Yektamaram", "email": "mr.yekta@yahoo.com", "user": {"login": "mohammad1ta"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d669fc78c3013b4557ac8efd6f1477c3bd8a6ac3", "messageHeadline": "Revert \"use MLAS for nuphar's pool ops (#1914)\" (#1933)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 125, "deletions": 417}}, {"node": {"oid": "80ef629c02fd18393ceb3c26dd48a546635a314c", "messageHeadline": "Add CumSum and Round for Opset 11 (#1705)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 609, "deletions": 18}}, {"node": {"oid": "e6ce384402de43f6692f860d2d5cfbb2a98f5c98", "messageHeadline": "add dependency 'cub' as submodule (#1924)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 43, "deletions": 1}}, {"node": {"oid": "9ed85987e3cc4c76cdbbec0f1aa84be7660adb43", "messageHeadline": "ReduceMax/Min performance improvements on CPU (#1925)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 24, "deletions": 7}}, {"node": {"oid": "c9b9e79e382ffa650974291de22d3a0a97ace977", "messageHeadline": "logging related fixes. (#1930)", "author": {"name": "George Wu", "email": "George.Wu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "bd2d6af9ca47f21880fe121d3a7c5392f77d9826", "messageHeadline": "Filter out info from non-const initializers during shape inferencing \u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 227, "deletions": 87}}, {"node": {"oid": "8c809dcc9937a683075cce314abb3df08a610614", "messageHeadline": "use MLAS for nuphar's pool ops (#1914)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 417, "deletions": 125}}, {"node": {"oid": "d46e023ee45c9657dfa75320ebc98d0bff138a66", "messageHeadline": "Remove toolset=14.11 for CUDA build (#1921)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "052339d9dc553e8d3e39100575ca30dac8bbb419", "messageHeadline": "Fix python packaging pipeline (#1922)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "76f76251fad1c826d70d2dc66b9844bd691cac10", "messageHeadline": "Rewrite the softmax op (#1911)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 110, "deletions": 295}}, {"node": {"oid": "4d26f2ce86edeb27e0d9c69d83d48e73f864fd81", "messageHeadline": "Address two issues: (#1905)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 63, "deletions": 17}}, {"node": {"oid": "30c7c7655295cee582304462057485874af98930", "messageHeadline": "fix the output size param's location in the csharp OrtRun interop cal\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "baf75c44e721e48da3079df0bad0c5c774033a73", "messageHeadline": "Remove registration of Aten op. (#1908)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 0, "deletions": 38}}, {"node": {"oid": "e769617f75ef2ecd68cc6186998b0f6ba3280597", "messageHeadline": "extract pooling attributes into a separate class (#1906)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 253, "deletions": 226}}, {"node": {"oid": "0317e825f1b81b9e37ba917ea1a7929a1742fba7", "messageHeadline": "onnxruntime_perf_test: Add -y support to control parallel executor #t\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dbff8272e7e5a982a2d9e31e4067f3a6a82a101d", "messageHeadline": "Update ONNX to newer commit (#1907)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "28a62f772859d1d2ce43c99072dddec76a7b6adf", "messageHeadline": "MLAS: add U8S8 MatMul operation (#1895)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 5418, "deletions": 1575}}, {"node": {"oid": "d3cb2a5572b879441debcaac99da86d706f28869", "messageHeadline": "Add ScatterElements CPU kernel  (#1796)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 114, "deletions": 34}}, {"node": {"oid": "034aa8016767db946887d2ae53b05ddee76592d3", "messageHeadline": "InferenceSession ctor with byte array in C# (#1883)", "author": {"name": "yeohan", "email": "35736802+yeohan@users.noreply.github.com", "user": {"login": "yeohan"}}, "additions": 76, "deletions": 6}}, {"node": {"oid": "294db0f978a57e3f0e6999e12e40421563505d98", "messageHeadline": "Implement Determinant operator. (#1880)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 136, "deletions": 0}}, {"node": {"oid": "aa92daecc56c6a93edfa5c2419bda6ea2277d5d7", "messageHeadline": "Revert \"A small improvement to the parallel_for  when task count > th\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 17}}, {"node": {"oid": "686bd36210affca89d3791e696e84943a57d2f7b", "messageHeadline": "Remove ml_status.h, add StatusCode to pybind exception mappings (#1889)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 213, "deletions": 184}}, {"node": {"oid": "77176e8678b95ddb3f8de18c443bcb22d10fb9ae", "messageHeadline": "Fix broken link to mnist model (#1896)", "author": {"name": "Louis-Philippe Bourret", "email": "lobourre@microsoft.com", "user": {"login": "LPBourret"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3d73a69edaca17569c0bdfcc0ffceadad5f638d9", "messageHeadline": "Stablize Nuphar test (#1894)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e288b871eafeca5e57633f5fc0eeee8d87a5848a", "messageHeadline": "CUDA Equal Greater Less can't support multi-directional broadcast", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 196, "deletions": 87}}, {"node": {"oid": "1a3ded6a7b15f867a18a8f501909b852e7eaa13a", "messageHeadline": "Add C API for free dim override, fix missing API mention in Inference\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 28, "deletions": 7}}, {"node": {"oid": "49a4233bf337ae354e1402c1e8f8078d9017c2a4", "messageHeadline": "Account for Constant node being versioned in opset-11 (#1875)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "ca89387817ef83849f51f93056cadaa7aba05534", "messageHeadline": "Build OnnxRT with Openvino EP on Windows (#1865)", "author": {"name": "avidiyal", "email": "41489828+avidiyal@users.noreply.github.com", "user": {"login": "avidiyal"}}, "additions": 86, "deletions": 39}}, {"node": {"oid": "1f4190de3c4a5f4f85df76bf6df097e72dbaf1e2", "messageHeadline": "Disable Linux x86 builds as they're not required any more. (#1887)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 0, "deletions": 97}}, {"node": {"oid": "aacfa2af65080399243c24a31c88f94ca5703575", "messageHeadline": "Bump up ONNX to the latest commit (#1868)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 364, "deletions": 259}}, {"node": {"oid": "9707b39a67cbe34a00d8a9917eea3577d66e70c1", "messageHeadline": "Duli/clip cuda (#1677)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 124, "deletions": 0}}, {"node": {"oid": "57812224560954dcc1719fb0f28d318235db5f5f", "messageHeadline": "Ryanunderhill/api interface (#1855)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1479, "deletions": 1089}}, {"node": {"oid": "a7beed798e8bf25352cf315e24ad784f6ef90538", "messageHeadline": "Implement L1 graph transformer for free dimension override (#1825)", "author": {"name": "Adrian Tsai", "email": "adtsai@microsoft.com", "user": {"login": "adtsai"}}, "additions": 213, "deletions": 3}}, {"node": {"oid": "561f2c4a9aa6e2c71ad817ba40891fe364b4545a", "messageHeadline": "Update pybind (#1876)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "6a9ae65f4199b124d03db49e8923969e798fab68", "messageHeadline": "Expose GetOverridableInitializers via Python and C/C++ API (#1878)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 237, "deletions": 35}}, {"node": {"oid": "429f05138aad4ce756ba13e1f42b5c91811b450f", "messageHeadline": "[OpenVINO-EP] Disabled div unit tests for GPU (#1853)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 33, "deletions": 9}}, {"node": {"oid": "60e5eee52afe0865c2e476f7263ac885f47b2fa3", "messageHeadline": "Improve Session Capabilities in ORTServer (#1862)", "author": {"name": "Colin Versteeg", "email": "cverstee@gmail.com", "user": {"login": "csteegz"}}, "additions": 103, "deletions": 32}}, {"node": {"oid": "b43254282fb0699992bcfca63ceed7f74168a663", "messageHeadline": "Handle bad alloc exception in bfc arena (#1846)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 23, "deletions": 15}}, {"node": {"oid": "a9ce9415796b9ab9c6ed9631e714b8067b393b50", "messageHeadline": "Refine threading control options and move inter op thread pool to ses\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 286, "deletions": 200}}, {"node": {"oid": "938b3627b802d0cd0b19362338c88bf542cca4b2", "messageHeadline": "Improve error message if node throws (#1847)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 142, "deletions": 96}}, {"node": {"oid": "6792e00e1ef77eb33b4765514082f42406ae4697", "messageHeadline": "Revert accidental TVM change (#1874)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "55f5a52b23ee5e5c19d1597e9304383f24c95a0a", "messageHeadline": "Support opset-11 DepthToSpace CPU kernel (#1759)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 166, "deletions": 18}}, {"node": {"oid": "80bda772037717856d15fb05d3dbd3a2090d177a", "messageHeadline": "Fix symbolic shape inference for faster_rcnn, mask_rcnn, yolov3 (#1867)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 33, "deletions": 11}}, {"node": {"oid": "582a27f5461c6fb6c68f626c77edbd3ae4711ba0", "messageHeadline": "remove sudo from the cleanup step for Linux so that we don't need the\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "6e4e76414639f50836a64546603c8957227857b0", "messageHeadline": "upgraded CSharp test and sample projects to netcoreapp2.1 (#1869)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 614, "deletions": 300}}, {"node": {"oid": "c0d953a26821396a6b6b07006fc5a93c78c69cd0", "messageHeadline": "Support opset-11 Loop CPU kernel (#1816)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 98, "deletions": 4}}, {"node": {"oid": "dc03ce0278d188ebdaf05c5dcd8995412ab75c43", "messageHeadline": "New OP: CDist (#1808)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 372, "deletions": 27}}, {"node": {"oid": "d646dc73e45aa0fa36db679e92197a298198e627", "messageHeadline": "OpSet 11 Operator Update: Supporting Neg Axis (#1835)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 191, "deletions": 38}}, {"node": {"oid": "2f7f83c655a209ae8361b76de157a961a30a954e", "messageHeadline": "ROIAlign: add cuda kernel implementation (#1823)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 399, "deletions": 20}}, {"node": {"oid": "71d414cacb0c49b797ae3cc3cb0af54a71e8999c", "messageHeadline": "update tvm submodule to the latest (#1849)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "fc1c9971a30878d50c478688338d70034502d56f", "messageHeadline": "Support replacing OrtValue of feed in IOBinding instance (#1819)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 46, "deletions": 23}}, {"node": {"oid": "166b1f86db0d9a41189a65b1bfb30110424cc955", "messageHeadline": "A small improvement to the parallel_for  when task count > thread cou\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 17, "deletions": 3}}, {"node": {"oid": "8a9c4cd9368c2157a3a4f1e59c834ff3d0fa3466", "messageHeadline": "ROIAlign: fix bilinear interpolate in 'max' mode (#1821)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 13, "deletions": 11}}, {"node": {"oid": "a1eecd8087af24b0037eb11668ff3368bb0db818", "messageHeadline": "Fix C# build (#1834)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "125900c96156ecf6bd421d3824bd2f66007fc3d9", "messageHeadline": "Enable integration with mimalloc memory allocator (#1673)", "author": {"name": "kile0", "email": "kile@microsoft.com", "user": {"login": "kile0"}}, "additions": 95, "deletions": 3}}, {"node": {"oid": "bd5451b4ed9c86f30556ed6514bf0bc4c1948a76", "messageHeadline": "Don't define USE_OPENMP if the compiler doesn't support OpenMP (#1836)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 16, "deletions": 17}}, {"node": {"oid": "aac60215492742322fae85732a87d3e870953536", "messageHeadline": "Add NuGet feed publish to nuget pipeline (#1833)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "2a21df23095efaad89c8de1d447f3a53b1219d31", "messageHeadline": "Updates to CUDA and TensorRT dockerfiles for v0.5.0 (#1731)", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 45, "deletions": 34}}, {"node": {"oid": "634f554471b5b838dcc52e0a00237a27fb718620", "messageHeadline": "python api's for execution provider registration (#1826)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 267, "deletions": 149}}, {"node": {"oid": "cf22ea68935c21c5beac217eb816f82940e65f7f", "messageHeadline": "Upgrade TVM for a fix in tvm::Integer with int64_t input (#1824)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a60283845b6e3a837c828f2d4c904aa0edbf83a5", "messageHeadline": "Update link format and example sections in readme (#1729)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 68, "deletions": 55}}, {"node": {"oid": "a0ba25f98f210fa506300bb5040695e2b7a636a8", "messageHeadline": "Fix the issue that it run into alloc failed on multiple cuda device. \u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 84, "deletions": 88}}, {"node": {"oid": "18f737726948591097495a1969599e0789f033ce", "messageHeadline": "[Nuphar] improvements in symbolic_shape_infer and model editor (#1787)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 487, "deletions": 215}}, {"node": {"oid": "8712a523a4cda4500256aafae378558e64d1f672", "messageHeadline": "Bump onnx to latest (#1756)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 265, "deletions": 49}}, {"node": {"oid": "f8c34428805909864f273fe7ac287bc2fe079a9b", "messageHeadline": "Part 2 of renaming AllocatorInfo to MemoryInfo. (#1804)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 180, "deletions": 180}}, {"node": {"oid": "397713a2d12ebd968a9cd0cd436ec9e55047922a", "messageHeadline": "Add Int64 support to ReduceMin, and Add Double support to Neg op (#1812)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "2b8677b2102b47a75c85422d4e2db1385b5e49be", "messageHeadline": "Enable Openvino nightly build on edge device (#1684)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 100, "deletions": 34}}, {"node": {"oid": "fe8915863cdb2c39edfb41f5f3dccef6d2389ada", "messageHeadline": "Implement C API entry points for creating and fetching non-standard t\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 794, "deletions": 72}}, {"node": {"oid": "d9fa632863698c083a06004c5de8ea216ba35c7b", "messageHeadline": "Add Cuda Kernel for Not operator (#1801)", "author": {"name": "Chi Lo", "email": "54722500+chilo-ms@users.noreply.github.com", "user": {"login": "chilo-ms"}}, "additions": 28, "deletions": 1}}, {"node": {"oid": "a9e4de2ceabdf9b72cdd890144c12734987cae18", "messageHeadline": "Follow up on proto3 compatiblity. (#1799)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 43, "deletions": 17}}, {"node": {"oid": "3b7f047a49576a4d9b83c6089079cc040c2dd1d8", "messageHeadline": "General performance testing tooling improvements (#1577)", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 116, "deletions": 50}}, {"node": {"oid": "6586afc8eb16abdbdb383d4752405721653f541b", "messageHeadline": "Refine the output shape calculation to avoid unnecessary re-allocatio\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 15, "deletions": 5}}, {"node": {"oid": "35c5c4d418403b30dbce2ad96caeff9141bca376", "messageHeadline": "A subgraph may have no inputs (e.g. subgraph in If has no explicit in\u2026", "author": {"name": "Scott McKay", "email": "skottmckay@gmail.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "206278ca44e7e50869c60676e84710cdbc508c9f", "messageHeadline": "Fix error message in Cast op (#1792)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f9d85d654a7f308386514db68ccd9346cc291cad", "messageHeadline": "Add GetDataTransfer() interface in the EP. (#1773)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 40, "deletions": 18}}, {"node": {"oid": "bd486605926df1d95597872327947562cf737981", "messageHeadline": "Add Cuda Kernel for Less operator (#1790)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 72, "deletions": 1}}, {"node": {"oid": "b32f24a3f946b3732d9432d68838f98530b9f5f6", "messageHeadline": "added support for Less(double) (#1722)", "author": {"name": "Ran Cohen", "email": "ran.a.cohen@intel.com", "user": {"login": "RanACohen"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "0b609d3e682f979887b31b6b8d4137e03bb79977", "messageHeadline": "Add make_unique implementation for use with C++11. (#1793)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 615, "deletions": 248}}, {"node": {"oid": "98dbdb1e0bb4b1a03013503b78d9a964816bbf83", "messageHeadline": "Rework the feed/fetch copy setup so that it can be calculated prior t\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 1057, "deletions": 847}}, {"node": {"oid": "2e242a40898e0b2b4ccaf63a21a9f71ffd06954a", "messageHeadline": "Clarify naming of the API involving the RunOptions terminate flag. (#\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 22, "deletions": 19}}, {"node": {"oid": "75f241d02c4cf6b35b326d72267364948bc4f8af", "messageHeadline": "Enhance compatibility with proto3 and replace or abstract has_*() met\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 431, "deletions": 273}}, {"node": {"oid": "6a5b11756b85c3324f4bae47be3dcb61167a6a06", "messageHeadline": "Conditionally export execution provider apis in chsarp (#1724)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 194, "deletions": 132}}, {"node": {"oid": "071a0c2522dd2af2b1936d608ef7182be6cfa883", "messageHeadline": "MLAS: MlasSgemm refactoring (#1749)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 3332, "deletions": 3630}}, {"node": {"oid": "a324ad7b96a831c5ed83e873d9429da05bf887be", "messageHeadline": "MLAS: clang u8u8 GEMM fix", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "b2a2326a454b00aa879f8a19ca278c0457b3a5b7", "messageHeadline": "add dequantize and quantize back to contrib ops (#1712)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 452, "deletions": 32}}, {"node": {"oid": "e1a12b17609b5eeb0d6316539b8dcb66db34bc48", "messageHeadline": "Fix some unnecessary copies of the Node attributes (#1763)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 9, "deletions": 8}}, {"node": {"oid": "52fe574fedf001d879007b165c5a5ea251b2976c", "messageHeadline": "Rename OrtAllocatorInfo to OrtMemoryInfo to make it more obvious. (#1\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 215, "deletions": 219}}, {"node": {"oid": "58fe5a6bf1fc916964ae5036df35c038ff9b21c6", "messageHeadline": "Enable Nuphar docker build, and reinstate Nuphar tests (#1757)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 74, "deletions": 28}}, {"node": {"oid": "eddb9d78f9d790d2fa82916a597be941a2f3952e", "messageHeadline": "fixed \"unreachable code\" warnings on Windows (#1755)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "7c5b3a5eccaeab7c3dea87ca17d09e64ffcb421e", "messageHeadline": "Update coding guidelines to prefer using make_unique for heap allocat\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "3d44c550921c5a76d69af74b50d9d377a82d9013", "messageHeadline": "Updated docs related to base images (#1753)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "4ed8d4b30e1f176b51ffcb4ab8d44761fb4534d2", "messageHeadline": "Put the initializers at the end of the cluster inputs list (#1751)", "author": {"name": "Tomasz Do\u0142bniak", "email": "tomasz.dolbniak@intel.com", "user": null}, "additions": 13, "deletions": 2}}, {"node": {"oid": "9523977cc2bbd0fa30219f938bad273dbe5bb507", "messageHeadline": "Added emotion ferplus support (#1752)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "94d9161166fb820ed1e475b4198a5c26783f3e46", "messageHeadline": "Add nuphar to Linux CI build (#1750)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 32, "deletions": 11}}, {"node": {"oid": "0f6cf9a335d4b872644b5455f2f4fba64cc6b598", "messageHeadline": "enable quantizing specific nodes (#1742)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 45, "deletions": 12}}, {"node": {"oid": "ad7ab3d880311a60b3b47f3876c57f38c84a367a", "messageHeadline": "Enforce shape validation. (#1716)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "c9240f4e932098665d356387fc82502e1aa8180c", "messageHeadline": "Implementation of Nuphar execution provider (#881)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 15443, "deletions": 147}}, {"node": {"oid": "f4a6d267c18e20a7222d8e282140962770a4fbfe", "messageHeadline": "MKL-DNN EP:  control flow fix (#1740)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 6, "deletions": 8}}, {"node": {"oid": "259863758e19a2b9752a165a1113b9879048df49", "messageHeadline": "Fix typo in NMS code", "author": {"name": "Takeshi Watanabe", "email": "take-cheeze@users.noreply.github.com", "user": {"login": "take-cheeze"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "dc9c89546d68f1be4745694f9834d968801f6839", "messageHeadline": "Update the docker file for OpenVINO (#1741)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 28, "deletions": 29}}, {"node": {"oid": "833e18345d88a6ff2d72749e8b8e4018f221190c", "messageHeadline": "Publish perf tool with nightly build (#1728)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 24, "deletions": 0}}, {"node": {"oid": "810ee0068f5d9c91426fe941f961410792337b7a", "messageHeadline": "Fix a issue that CUDA EP fallback to much nodes to CPU for some case \u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "25d02a33c814e8bf2375f8b39d0d9b4fdcf3993a", "messageHeadline": "Fix reading of onnx domain causing one of the automl models to break \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 42, "deletions": 6}}, {"node": {"oid": "e54904e6a3684c329a248993ff577901255a20ca", "messageHeadline": "add implementation for dynamic quantize linear (#1697)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 160, "deletions": 17}}, {"node": {"oid": "4b5b0372899b0685eb5f08db2192faca924e12f0", "messageHeadline": "Support 'Bilinear' mode for 2D inputs in Resize and Upsample kernels \u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 270, "deletions": 52}}, {"node": {"oid": "0f7c01b49b2c64dc7b66a654e4320eb2c536abfa", "messageHeadline": "Use exec form of ENTRYPOINT for docker server (#1690)", "author": {"name": "rakelkar", "email": "rakelkar@gmail.com", "user": {"login": "rakelkar"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "068b568472bab389aa6372c27a752638f5a26eb3", "messageHeadline": "Add support for int8 x uint8 for MatMulInteger, and int16 x int16 cus\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 274, "deletions": 12}}, {"node": {"oid": "8fc8910a0ed7461524432094d1b6dd144c6d9eeb", "messageHeadline": "Allow input used across execution providers as long as they use the s\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 6, "deletions": 6}}, {"node": {"oid": "81ad48080b027b9e91e2d42186f0b89a982377a1", "messageHeadline": "Remove TaskThreadPool (#1713)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 64, "deletions": 384}}, {"node": {"oid": "73312b8195e686163a11ca5a15fab7384bf9d841", "messageHeadline": "MLAS: Android sgemm kernel build fix (#1710)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 28, "deletions": 53}}, {"node": {"oid": "14eae293bf84956f371bfac50628e15b73751940", "messageHeadline": "remove @PCGOTREL x64 usage (#1707)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 48, "deletions": 46}}, {"node": {"oid": "d9cdf4b4ed7a4058beaa1ce14ae7c2521beddf0a", "messageHeadline": "Doc updates (#1522)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 257, "deletions": 185}}, {"node": {"oid": "8813b79c5be90c41c9dbb7d4d4244500f13db4e0", "messageHeadline": "make gemmlowp default for arm (#1701)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "121d308a33810729e8b5e9a63ed23799711889cf", "messageHeadline": "Python API naming and other cleanup (#1678)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 27, "deletions": 30}}, {"node": {"oid": "938200de9b800e436349ff64a57898d4139f932e", "messageHeadline": "fix typo in max batch size error msg. (#1687)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "961b14ac4a397532d5a76cd6f360aeb2074d2444", "messageHeadline": "use MLAS for QGEMM in matmulInteger and convInteger (#1692)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 123, "deletions": 40}}, {"node": {"oid": "a8998b07b57d35195518b7190858c970494027da", "messageHeadline": "treat zero point properly (#1686)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "f25847bccde222eacc3a5f8d515997d01631a428", "messageHeadline": "More fixes on the NuGet CPU CI pipeline (#1688)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "5873bdbb3f66fb761e4846893579c803b608eba2", "messageHeadline": "Share default CPU allocator with Mlas preferred alignment (#1682)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 34, "deletions": 22}}, {"node": {"oid": "4035fe842e7d18e43b27f660a7c2c992220e8e60", "messageHeadline": "Don't create the default allocator every single time. Rename API acco\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 68, "deletions": 60}}, {"node": {"oid": "7408dec0bfee82c433bb26856dd204dfc9ec90a2", "messageHeadline": "Added some mo optimizations to improve performance (#1674)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 31, "deletions": 5}}, {"node": {"oid": "addf32fa2a55279bd3f6f46c258f84aa9ef0c56b", "messageHeadline": "int64 support for 'where' op (#1666)", "author": {"name": "Negin Raoof", "email": "neginmr@utexas.edu", "user": {"login": "neginraoof"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "c9a4fe2b7b63ce173b5574a81546f3ef802a5ad0", "messageHeadline": "Add support of ReduceSum int64 (#1664)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 22, "deletions": 1}}, {"node": {"oid": "d2569d37615a7e46bf4f33547f00be100eff39d6", "messageHeadline": "update clip for opset 11 (#1661)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 97, "deletions": 15}}, {"node": {"oid": "4de0aa8049f9967d4e91020286f383022aa0b1fa", "messageHeadline": "Optimize kernel index (#1672)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 214, "deletions": 483}}, {"node": {"oid": "a818740d914e90722fa8936e939a6c88b00fc72d", "messageHeadline": "Support Tensor<bool> and Tensor<Int8> in C# API. Support Tensor<strin\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 146, "deletions": 38}}, {"node": {"oid": "b53f40a886114bbcd093fe3a417b4c233fb3ed2c", "messageHeadline": "update set fetches for execution with allocation plan. (#1668)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 8, "deletions": 14}}, {"node": {"oid": "6f70a78e1f6a4808f6a479af601437a3d4442706", "messageHeadline": "Fix a few errors in the NuGet pipeline (still broken) (#1656)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 20, "deletions": 11}}, {"node": {"oid": "97d0a46afc240d9f6dc3ef3b42f685db6b0252d8", "messageHeadline": "nGraph EP Optimizations (#1630)", "author": {"name": "Tommy Trimeloni", "email": "thomas.v.trimeloni@intel.com", "user": {"login": "tvtrimel"}}, "additions": 50, "deletions": 6}}, {"node": {"oid": "a68a20e4150f05ae5f572e8eec45c0329e447e54", "messageHeadline": "Add details of which node was not able to be placed on an execution p\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 6}}, {"node": {"oid": "e652a236b492e573bdaafe12b6fbe810de73aae8", "messageHeadline": "cudnnRNNForwardInferenceEx doesn't support 0 sequence in the bathes", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 192, "deletions": 6}}, {"node": {"oid": "d0d82432f3f37ab8609ec2cb4671cf3dd51f8484", "messageHeadline": "Update PyTorch Section for supported onnx version (#1635)", "author": {"name": "Emma Ning", "email": "43255631+EmmaNingMS@users.noreply.github.com", "user": {"login": "EmmaNingMS"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5311c1b2b5126dffebc1e1a90c531870c901666b", "messageHeadline": "Check return value form CreateFeedsFetchesManager. (#1653)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "7be5695fad669d58c104fc5c0db12a391e00f0e2", "messageHeadline": "Remove --whole-archive (#1655)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 32, "deletions": 29}}, {"node": {"oid": "68d496c7ca6e10437967f234541b0b2818b6e670", "messageHeadline": "fix bug on windows where ops were always getting dumped. (#1648)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a1b3c64038779df617c2e2661583e5b9c7346b75", "messageHeadline": "Fix memory leak in mlas unitest (#1654)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "377dcf60ac17aa1985ab404b63b227d1ae5d3d10", "messageHeadline": "Update onnx test runner documentation (#1651)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 14, "deletions": 11}}, {"node": {"oid": "224dde7ef119959ffc2d18420d7b9e76121faa19", "messageHeadline": "Allow user disable multiple threading (#1647)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 15, "deletions": 15}}, {"node": {"oid": "6f3a835d389ecdec7c0d1af09153156309e31d45", "messageHeadline": "Update perf tool documentation to reflect the new graph optimization \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 9, "deletions": 4}}, {"node": {"oid": "413730365fed93036635ecc71be9c9b6663ca8ed", "messageHeadline": "MlasGetMaximumThreadCount: plus 1 to the NumThreads from ORT thread p\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "bdc694314cbca0f5931333579efa54b8e07f3175", "messageHeadline": "update MKLML to version which contains fix for thread hang. (#1636)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "b963e4b5c220cda8740b85f4b09064c57403651d", "messageHeadline": "Add uint8 Support for NonZero Op (#1614)", "author": {"name": "Lara Haidar", "email": "haidar.lara@gmail.com", "user": {"login": "lara-hdr"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "bc72c2dba792c7b58c90b776854ee78ff3ee8369", "messageHeadline": "MLAS: add U8U8 MatMul operation (#1644)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 4958, "deletions": 59}}, {"node": {"oid": "fbd790f70323f39885aa7838fb047fa4ac1bba75", "messageHeadline": "Add AutoML to 3 main builds. (#1631)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 125, "deletions": 135}}, {"node": {"oid": "372b657900a402a38067c70ab456c8e5ca8dfe6e", "messageHeadline": "update TRT EP CI's to use latest model.zip (#1637)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 17, "deletions": 7}}, {"node": {"oid": "6b89c7ad0416ab63e39d0900b363968adb30d04c", "messageHeadline": "Let mlas use session thread pool (#1609)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 450, "deletions": 446}}, {"node": {"oid": "44a42a6a9835f2b84adb140cba6b7dc887c719d1", "messageHeadline": "Fix parsing initial hidden state in RNN (#1626)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 59, "deletions": 3}}, {"node": {"oid": "f9834105aa07f47d277328312ac240bb0a992c37", "messageHeadline": "removed --gen_doc (#1633)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c9eb13a638c1de0bf0784b0cb3208d1198d59883", "messageHeadline": "Copy System.Numerics.Tensors sources from dotnet/corefx into onnxrunt\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 22257, "deletions": 84}}, {"node": {"oid": "0044be62590a2a9f52148c115a4f3905d3cb12bc", "messageHeadline": "update onnx to latest commit (#1622)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 55, "deletions": 9}}, {"node": {"oid": "1835640d943eb8239f10fdffb803f1f7fcdd9732", "messageHeadline": "Support int64 for ReduceMax (#1625)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 28, "deletions": 0}}, {"node": {"oid": "17c8fe44e37bfbe007a9f36fd105fb29a26181e3", "messageHeadline": "Integrate featurizers (#1573)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1607, "deletions": 6}}, {"node": {"oid": "7545b795dfcea271e8a964354d4d012c8c7bdd80", "messageHeadline": "Fix incorrect box offset computation in NMS op (#1624)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 37, "deletions": 2}}, {"node": {"oid": "0c5d2c998bb2f75551c4336f577d9d4b9efa3918", "messageHeadline": "Generate documentation from the registered operator kernels (#1395)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 829, "deletions": 33}}, {"node": {"oid": "8d12ce45cfc49f942c27639b8bbb437a0cea80a7", "messageHeadline": "Use a friendly enum for graph optimization level. (#1586)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 226, "deletions": 142}}, {"node": {"oid": "24d17f43533c0e15ec267d52c6b2f995aa5de65b", "messageHeadline": "Fix trtlogger segfault. re-enable SoftPlus unit test for TRT. add doc\u2026", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 27, "deletions": 6}}, {"node": {"oid": "09db1e06b5f94827af00709665e57c4958b2aa7a", "messageHeadline": "Make changes to pipeline template to include missing headers in tars/\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "a6a5acedda4e2b5190d6a7a9ea707c0b719c7ca0", "messageHeadline": "Cleanup csharp API SessionOptions and RunOptions to be consistent wit\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 602, "deletions": 179}}, {"node": {"oid": "bd64ca3019a470c519f9beaedcfaea73f84b4c37", "messageHeadline": "Kezhan/execute graph refactoring (#1553)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 97, "deletions": 111}}, {"node": {"oid": "b405482cfa3d38549273ffe47cfb56b4657a6a39", "messageHeadline": "Remove copy of generator in Multinomial (#1611)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 121, "deletions": 119}}, {"node": {"oid": "b5de1324ef881a54aee7e358ccf10701292e4574", "messageHeadline": "Fix log message truncation on Windows when printf formatting is used.\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 46, "deletions": 6}}, {"node": {"oid": "a50a63aa9ee501bbf8236ab44a07d7e53d2cb12b", "messageHeadline": "Serialize optimized onnx model (#1470)", "author": {"name": "pulkittomar", "email": "pulkit.tomar@outlook.com", "user": {"login": "pulkittomar"}}, "additions": 150, "deletions": 1}}, {"node": {"oid": "8a559d75ae4360f65aa5f71c1a698a7ea496e784", "messageHeadline": "Minor perf improvements. (#1580)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 45, "deletions": 26}}, {"node": {"oid": "a6a4c4c079ef6f84d3e2b90a9d62e4c5b66a6296", "messageHeadline": "Fix perf test executable. (#1598)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 351, "deletions": 420}}, {"node": {"oid": "ce3c8f98ddde318812a656ad63be082b6ddc342f", "messageHeadline": "Fix for CPU random ops seed narrowing conversion. (#1594)", "author": {"name": "AlbertSadovnikov", "email": "Albert.Sadovnikov@gmail.com", "user": {"login": "AlbertSadovnikov"}}, "additions": 35, "deletions": 20}}, {"node": {"oid": "df9b1b8ec83ebdae0026f7029a14d049aa7b860e", "messageHeadline": "Include io_win32.h only if builds on windows (#1587)", "author": {"name": "Malik Shahzad Muzaffar", "email": "shahzad.malik.muzaffar@cern.ch", "user": {"login": "smuzaffar"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "69baf9e8004b3220fb68fd9716a9a1aad6002b78", "messageHeadline": "Update nGraph to v0.22.1 (#1582)", "author": {"name": "Tomasz Do\u0142bniak", "email": "tomasz.dolbniak@intel.com", "user": null}, "additions": 218, "deletions": 270}}, {"node": {"oid": "7be40b2946b1a866baa0db1c09bd3edb48c12ae0", "messageHeadline": "put all gemmlowp common code in one place (#1590)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 248, "deletions": 278}}, {"node": {"oid": "59c9d83f357083688d5d55fd2dde90bf32c3ca04", "messageHeadline": "add int64 support for less op. (#1604)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "0187d876cb402583e8da9bb154abb42cdee7242d", "messageHeadline": "Implement new LabelEncoder in opset 2 in ML domain (#1393)", "author": {"name": "Wei-Sheng Chin", "email": "wschin@outlook.com", "user": {"login": "wschin"}}, "additions": 311, "deletions": 5}}, {"node": {"oid": "6d783e8a07888a901e580f2243dc9694d8ea95cd", "messageHeadline": "Added license files in the base image (#1595)", "author": {"name": "manashgoswami", "email": "magoswam@microsoft.com", "user": {"login": "manashgoswami"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "9b83545f664be4f81451c9a96ca7748b4fa8c553", "messageHeadline": "Optimize Fence checking performance (#1593)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 139, "deletions": 69}}, {"node": {"oid": "1c5b15c2b89432f03877c8845f2d854387543aea", "messageHeadline": "Remove memory copy between TensorRT and CUDA (#1561)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 104, "deletions": 159}}, {"node": {"oid": "38d78542c32413bd0b88c8d252d09ec7496f62b6", "messageHeadline": "Fix race condition issue in RNN/LSTM/GRU (#1544)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 94, "deletions": 93}}, {"node": {"oid": "6e430c0526dbf95d9da5c4da737dc78c6a2fbd6d", "messageHeadline": "A few performance improvements coming out of ssd_mobilenet and ssd_re\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 113, "deletions": 90}}, {"node": {"oid": "a443b013dd4f3e31c0f27b8e5d68e416140ab507", "messageHeadline": "Remove unneeded C APIs + some refactoring. (#1555)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 779, "deletions": 190}}, {"node": {"oid": "a93ece27274e6264bbcf36522a6ca696df68a578", "messageHeadline": "update quatizelinear to process int8 input (#1576)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 36, "deletions": 11}}, {"node": {"oid": "aeb0bcb4a3d838ddf215ad27a96bb8b3f9935a17", "messageHeadline": "parallel build", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "9a34089f67ac6e3c76e16e16f6a3409350d0c15d", "messageHeadline": "Add more type support for OneHot op (#1565)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 34, "deletions": 0}}, {"node": {"oid": "9e926fef1cf7c0ed9ccde5e88453fd91266f3c9b", "messageHeadline": " Add a doc for cmake (#1524)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 83, "deletions": 0}}, {"node": {"oid": "65ff02fdb0139a8ef13e0babd463439cb98c743d", "messageHeadline": "  Set job timeout for code coverage pipeline to 120min(#1563)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "16087f313360d81d8a3c208101f7b5297d0c4315", "messageHeadline": "update default values for weight quatization (#1564)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "7ee8aca1bff1eae633eacf16b4526171ae13f149", "messageHeadline": "Avoid downloading test data into C:\\ (#1562)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 105, "deletions": 74}}, {"node": {"oid": "05bbb3065c40c7b0334fd9614f0cef262c057cd8", "messageHeadline": "[OpenVINO-EP] Update hardware branding of VAD-R as VAD-M (#1552)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 23, "deletions": 23}}, {"node": {"oid": "ceb8f1c1a257bf43615e94c272fa6e21e68df18f", "messageHeadline": "Modify the kernel declaration for Shrink op (#1554)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "6c271c63ac98c05ddf5c017a419fab1ef29cd7a7", "messageHeadline": "add test cases for commit c019bb9355a511f471e55e7302b26e1d370ed46a (#\u2026", "author": {"name": "pengwa", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 71, "deletions": 63}}, {"node": {"oid": "8a6bfe00af959351bb6cbf80acf7065d8ad55e74", "messageHeadline": "roll back model test update for ngraph provider. (#1551)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "a098be12ba227f85b01a48701c4ff87e641fd50e", "messageHeadline": "Register kernel for Greater int64 (#1546)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 128, "deletions": 337}}, {"node": {"oid": "cb71c69d5ed8160fc96818ef0b287770d3e6ff9b", "messageHeadline": "checking execution provider logic updated. (#1547)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 7, "deletions": 15}}, {"node": {"oid": "93cb29f9583306cfa7676440d989dafc05f6383b", "messageHeadline": "[WIP] NNAPI EP Update (#1540)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 26, "deletions": 20}}, {"node": {"oid": "9fb8867a2439ca28814540f24af9e0e2431cc34d", "messageHeadline": "Don't create implicit input for outer scope value if there is a subgr\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 177, "deletions": 24}}, {"node": {"oid": "1cf5ebc4c5e3f02a9e094a6f1521b3d481a2b94b", "messageHeadline": "copyfromhost/copytohost are not needed for mkldnn ep (#1532)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 3, "deletions": 28}}, {"node": {"oid": "624411bb69ab9a9c412ad552bda0059a1307401d", "messageHeadline": "Upload correct ESRP signed package (#1531) (#1534)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 5, "deletions": 6}}, {"node": {"oid": "3045a5f88bcee5491e95f5937af96d292456a458", "messageHeadline": "Update test data (#1512)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "465b30e3ca5ff3bd78d587e223397fa836ec2911", "messageHeadline": "Bug fix for shape of optional output in Dropout op (#1507)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 69, "deletions": 6}}, {"node": {"oid": "57e24820893e33bd2021460a037b87c093b0e50e", "messageHeadline": "Fix a bug in Expand cuda op implementation. (#1528)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "b59936001455537eb6726f9cf4f3be5bdd42301a", "messageHeadline": "enable sse4.1 optimizations for gemmlowp (#1529)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 11, "deletions": 0}}, {"node": {"oid": "28a6f6b11b358e677f5f11aecc6f0a73b5eb2dfa", "messageHeadline": "Add back MacOS leg of the Python packaging job (#1523) (#1526)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 43, "deletions": 1}}, {"node": {"oid": "4d768b3a0fb2856f57fa770330408e6aecd4bcbe", "messageHeadline": "Fix inclusion of ARM binary in the release pkg (#1513) (#1521)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "fb5d0fc538c59e5fc79fc1a999e8f304b8ea6a94", "messageHeadline": "Publish nuget package to azure blob store (#1525)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 18, "deletions": 3}}, {"node": {"oid": "0b0e32909ad872addf6a4385d3bffd557c493bfd", "messageHeadline": "NCHWc: Enable Conv/Add fusion for stride=2 convolutions (#1518)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 103, "deletions": 13}}, {"node": {"oid": "14d46ee8902d49867618c850cbb427a1161102d1", "messageHeadline": "Init prev_Ht for zero length sequence to avoid valgrind warning. (#1516)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 23, "deletions": 15}}, {"node": {"oid": "fb7bdd177b70dae802044a1eb3576c00344d5b70", "messageHeadline": "Profiler-IsEnabled (#1503)", "author": {"name": "Jorgen Thelin", "email": "jthelin@microsoft.com", "user": {"login": "jthelin"}}, "additions": 29, "deletions": 16}}, {"node": {"oid": "a86486ab7fb3e718f88ee419509b395d2c01692d", "messageHeadline": "Post binary sizes to dashboard database (#1517)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 190, "deletions": 0}}, {"node": {"oid": "44ab301586cf304325f13fcbf461498c0484fab6", "messageHeadline": "More C API changes. (#1519)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 74, "deletions": 65}}, {"node": {"oid": "cf73f63cb93f651d3007d16bd2bc5f7765f7781e", "messageHeadline": "Enable float16 MatMul+Add -> GEMM fusion for performance boost (#1506)", "author": {"name": "Dwayne Robinson", "email": "fdwr@hotmail.com", "user": {"login": "fdwr"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "cf5a4b5856594ad94f8b669c5be6929d452cd70a", "messageHeadline": "remove the GetStream from cuda ep. (#1514)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 6, "deletions": 13}}, {"node": {"oid": "d6a30485be2015115b8abb60d8240af8f4e8a977", "messageHeadline": "Rename Tensor.Size() to Tensor.SizeInBytes() (#1502)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 17, "deletions": 17}}, {"node": {"oid": "6f538dc861cf60b32fa18fe27c2f91c9205af877", "messageHeadline": "Support missing optional attribute in Squeeze operator (#1505)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 28, "deletions": 8}}, {"node": {"oid": "717e764e8eab9fd50f3292f5d61900a494630415", "messageHeadline": "Move Class CudnnDropout to cudnn_common.h  (#1492)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 52, "deletions": 38}}, {"node": {"oid": "8589be69b238c2f1abc610023d525edcf118cabd", "messageHeadline": "Organized build instructions (#1504)", "author": {"name": "Emma Yu", "email": "42980371+emxuyu@users.noreply.github.com", "user": {"login": "emxuyu"}}, "additions": 19, "deletions": 8}}, {"node": {"oid": "33ae28ccb18efd2804c165a40ed6370a1ab44b4d", "messageHeadline": "Empty double quota `\"\"` is passed to `find_package(Thread)`, causing \u2026", "author": {"name": "xkszltl", "email": "xkszltl@users.noreply.github.com", "user": {"login": "xkszltl"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "be16b274fc75dac89f410251ccf23b0563ed09fe", "messageHeadline": "Upgrade mklml and set march with official option. (#1469)", "author": {"name": "xkszltl", "email": "xkszltl@users.noreply.github.com", "user": {"login": "xkszltl"}}, "additions": 9, "deletions": 20}}, {"node": {"oid": "1a115ed8cb93fe37a7f1d40cd21ea4a9c4fb294d", "messageHeadline": "Fix buffer overrun bug in CPU upsample op (#1501)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6df4bc2ebe8bda56ce306b118d62c83850fb7d3a", "messageHeadline": "Update scripts to access pipeline variables correctly (#1499)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 27, "deletions": 12}}, {"node": {"oid": "e0829b2b13d6ceff3e99a6acb1b0f26f6065a1de", "messageHeadline": "Revert the last changes on tree ensemble classifier (#1498)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 79, "deletions": 213}}, {"node": {"oid": "4ace393bea5003000d5bf3d3572210599796bedd", "messageHeadline": "Fix sign-compare warnings with gcc", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "6625eecd096f1e9d230168a1e3191c9f5c44534e", "messageHeadline": "Temp fix for a crash in fused graph (#1488)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "a7223ed801d894cdff5dc588e6ba689d057f7e07", "messageHeadline": "Fix android build (#1489)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f052966972d2d7fc0b09572ceb40079f4f7799f1", "messageHeadline": "Remove special casing of \"None\" as a dim_param (#1482)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "a8e3ff47fd61be92da391d3c16e6eed1ca41dfa5", "messageHeadline": "Add no scale check for resize and upsample (#1484)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 129, "deletions": 59}}, {"node": {"oid": "258ff06e42ddbcb7a228fb18de5e50be658302e3", "messageHeadline": "Revert \"publish nuget package to azure blob (#1309)\" (#1485)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 27}}, {"node": {"oid": "ec3c553501f0940f61943fb355c6b390a727e7e8", "messageHeadline": "NNAPI EP Update (#1483)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 163, "deletions": 115}}, {"node": {"oid": "c0f927c57cfd73e3f9d1733022b6bf6a94f8f6ab", "messageHeadline": "docker updated to support openvino R1.1 (#1475)", "author": {"name": "avidiyal", "email": "41489828+avidiyal@users.noreply.github.com", "user": {"login": "avidiyal"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "91d32c90609df25ec3413b4fd361b1aebd754528", "messageHeadline": "Add docs for the fns candy demo (#1479)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 103, "deletions": 26}}, {"node": {"oid": "a8e9e1878e94eb42d2895bcc984d16d8760c7bf7", "messageHeadline": "Reduce artifacts size (#1477)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 22, "deletions": 8}}, {"node": {"oid": "bb268657581719f0ff03cabfa30714cef98f24fa", "messageHeadline": "Optimize the resize and upsample (#1426)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 108, "deletions": 17}}, {"node": {"oid": "4aa4ca1502e002f4ef794b84c4561b1f58774edf", "messageHeadline": "Relax shape validation checks. Log a warning instead of returning an \u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "c5f2f0f15bba9537f7e027c81f379953201176c9", "messageHeadline": "Upgrade version number for ORT in preparation for release (#1468)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 20, "deletions": 8}}, {"node": {"oid": "160165016116c2e180399bd03a17d29cb999501f", "messageHeadline": "publish nuget package to azure blob (#1309)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 27, "deletions": 1}}, {"node": {"oid": "be02214a172bd813183f0452e63b6291180447cb", "messageHeadline": "Add a comment to onnxruntime_cxx_inline.h (#1466)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b41f6eef5221265d544aa195ad34ef30ea98fb52", "messageHeadline": "Jignparm/copy cuda extensions (#1462)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 17, "deletions": 0}}, {"node": {"oid": "818c023535cf854071121137ff1ea11697c9e771", "messageHeadline": "Add/correct missing SAL annotations + avoid using unsigned types (exc\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 93, "deletions": 87}}, {"node": {"oid": "387d4c72bb98984695ba645fe4738255e5bb6b95", "messageHeadline": "Strip invalid dim_param and dim_value values out. Allow re-use in eve\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 54, "deletions": 18}}, {"node": {"oid": "bbe92035c6002afd96b1f32563291581c9c81767", "messageHeadline": "OStreamSink produces interleaved output due to issuing multiple opera\u2026", "author": {"name": "Jesse Benson", "email": "jessebenson@users.noreply.github.com", "user": {"login": "jessebenson"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "768ced703c49cf16545bc0324029b072f7261d0f", "messageHeadline": "Expose provider factory C API, especially for CUDA users (#1461)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 30, "deletions": 8}}, {"node": {"oid": "31838fc9ee26935d6ebc8607e9fef9fcfe567afb", "messageHeadline": "remove const_cast which makes it's not thread safe. (#1463)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "6be93f11e5d3418d758f2169e8d506971e22e103", "messageHeadline": "build mklml/ngraph without openmp (#1460)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 7, "deletions": 10}}, {"node": {"oid": "1f13a9f98207642dedc194ca63530a0dadbe68cb", "messageHeadline": "Update to include more samples (#1381)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 12, "deletions": 6}}, {"node": {"oid": "1fc6f8ee5bbc8fc39712ef7eb0847a6bf5c62502", "messageHeadline": "Support double type for a few ops (#1450)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 298, "deletions": 59}}, {"node": {"oid": "9d67292c8cc5c6dbe4b4975162cdc394ff37253c", "messageHeadline": "Document for the C/C++ samples (#1442)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 184, "deletions": 35}}, {"node": {"oid": "29de25c5a770ba94d863296c43613b915b7f7fe3", "messageHeadline": "Mention OrtCreateSessionFromArray in C API doc (#1459)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "a5f57f43c260536cbb3d2ae4806327d2f1d2f05e", "messageHeadline": "fix mkldnn linux build break. (#1458)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "f938a6e53a93d2071aaa30a06b4fea917214eee8", "messageHeadline": "Add test for LSTM/GRU which has shorter sequence in the middle (#1437)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 84, "deletions": 5}}, {"node": {"oid": "227734139a0628907ef94b79a6e80d46deb0b737", "messageHeadline": "Fix ORTSRV nightly build (#1440)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 9, "deletions": 11}}, {"node": {"oid": "ac25a2643b75701b3c3dec4272d56d32afffa83a", "messageHeadline": "add VS2019 CMake generator instrs (#1441)", "author": {"name": "kile0", "email": "kile@microsoft.com", "user": {"login": "kile0"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "2c052919081617f83d1929178947a8f5ca80d389", "messageHeadline": "Jignparm/patch 0001 (#1419)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1a957e0642c25fe94a97fbfbc2f2e7266caa036a", "messageHeadline": "Update C-API packaging pipeline to use CUDa 10 (#1445)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 2, "deletions": 17}}, {"node": {"oid": "9e4ac8c66afa947080036628bb3646f76e413d32", "messageHeadline": "remove mkldnn from gpu nuget package (#1443)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "df3a157dd1590bb195014d7632cc69dc4d7ac8c3", "messageHeadline": "Add noexcept to cxx api (#1448)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "995ec04c7b33e6f1d5d6418562e151329c31c08f", "messageHeadline": "add float support for equal op. (#1449)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 18, "deletions": 0}}, {"node": {"oid": "414a07a85b17be736739133473c1ab341d4b83e1", "messageHeadline": "Support Sum opset6 for ONNX 1.2 models (#1447)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 13, "deletions": 7}}, {"node": {"oid": "9e2fa697853c27114ad20a8d1322dff5319b0786", "messageHeadline": "Ryanunderhill/c api string arg (#1436)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 39, "deletions": 1}}, {"node": {"oid": "f0b9a814e5efe678d4b399d8e48f57d9fc031165", "messageHeadline": "fix return of flakey mnist test on gpu. (#1446)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 2}}, {"node": {"oid": "950b863e22ecfe8a4a36cf133e0eee8591fe0f8c", "messageHeadline": "Update ONNX Runtime Server documents for build and usage. (#1444)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 30, "deletions": 74}}, {"node": {"oid": "e5107fd0cbf2abaf1856d1ae38cc13b3b4ae4eda", "messageHeadline": "Support MultiD input data for OneHotEncoder op (#1343)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 38, "deletions": 12}}, {"node": {"oid": "751ee7bb23e63b54ab5a311053fedb0dca43331a", "messageHeadline": "Fix bug in TransformerMemcpy  (#1413)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 157, "deletions": 21}}, {"node": {"oid": "4cbc6e1cf5b58208e7e82c210ba820a77b914c54", "messageHeadline": "Validate input shapes. (#1352)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 176, "deletions": 96}}, {"node": {"oid": "638398e6757623d97c3d7e31a9fd64f3166ba44e", "messageHeadline": "sync onnx to get equal op with float support (#1432)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 12, "deletions": 6}}, {"node": {"oid": "7717ed71a9e29a0c7f927ff72a58b7a331bfe20c", "messageHeadline": "Fix for issue #1397, disable onnxruntime_PYBIND_EXPORT_OPSCHEMA is --\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "e9e777925f622ee0c5668e31daa79784fb1b994a", "messageHeadline": "[OpenVINO-EP] Added support for OpenVINO R1.1 (#1438)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 575, "deletions": 539}}, {"node": {"oid": "f3c74ec3e9ee7e80daa3ccc7e95c3249844e7e41", "messageHeadline": "Reduce memory footprint of MKL-DNN EP (#1429)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 139, "deletions": 13}}, {"node": {"oid": "887930e6c2b240379181398c6f94742c7f8f1096", "messageHeadline": "inference overheads optimizations (#1392)", "author": {"name": "Yulong Wang", "email": "yulongw@microsoft.com", "user": {"login": "fs-eire"}}, "additions": 23, "deletions": 24}}, {"node": {"oid": "07ecd59e8fafd82f04dd40481084c0108c199e78", "messageHeadline": "flatten conv2d when input_width==kernel_width (#1435)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 46, "deletions": 7}}, {"node": {"oid": "c843c393e44aec6a53cab664baff5716960160b8", "messageHeadline": "More code cleanup (#1406)", "author": {"name": "Yuan Yu", "email": "yuanbyu@gmail.com", "user": {"login": "yuanbyu"}}, "additions": 20, "deletions": 25}}, {"node": {"oid": "bbf64c2c45d7611e8448b1216ab70e5a73473888", "messageHeadline": "Update cgmanifest.json and ThirdPartyNotices.txt for DNNLibrary (#1431)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 513, "deletions": 0}}, {"node": {"oid": "7a681fb96409c554070c4669d39d17f8b2f37fba", "messageHeadline": "Improve build throughput and enable using the Visual Studio 2019 cmak\u2026", "author": {"name": "kile0", "email": "kile@microsoft.com", "user": {"login": "kile0"}}, "additions": 56, "deletions": 6}}, {"node": {"oid": "5ee0f185dcb906c3c16b02f09a97c6a47110a139", "messageHeadline": "Add GRPC support to ONNX Runtime Server (#1144)", "author": {"name": "Colin Versteeg", "email": "cverstee@gmail.com", "user": {"login": "csteegz"}}, "additions": 775, "deletions": 91}}, {"node": {"oid": "6c41809655338f63de1c7d97f2bbeadae7138fbf", "messageHeadline": "Build Shared Library with cuda 10.1 (#1418)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 6, "deletions": 4}}, {"node": {"oid": "02ded802ab0abe3fe4e1b0ac4c59e1d3eeacaffa", "messageHeadline": "cleanup more useless unique_ptr (#1427)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 79, "deletions": 34}}, {"node": {"oid": "1ff957f96e62e6eb4b4439469cd10648be748efc", "messageHeadline": "CUDNN_RNN_DATA_LAYOUT_SEQ_MAJOR_UNPACKED works with CUDNN_RNN_PADDED_\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 41, "deletions": 4}}, {"node": {"oid": "f720166887196b8a7e52cb9c5ef3fed48250e97e", "messageHeadline": "register gpu data transfer only when there's nvidia gpu related eps. \u2026", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 10, "deletions": 5}}, {"node": {"oid": "db61eb4cd7eb3365193394b7d3ae9e893b0da85e", "messageHeadline": "Update ONNX_Runtime_Perf_Tuning.md (#1378)", "author": {"name": "Chris Seymour", "email": "chris.j.seymour@hotmail.com", "user": {"login": "iiSeymour"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f47f6fd02046ff4d397fbc8b4619c0db31863a35", "messageHeadline": "Fix MaxPool when using dilation > 1 plus non-zero padding (#1320)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 147, "deletions": 61}}, {"node": {"oid": "fbdd9054406bed6f931c4d744db3d2b1920079e6", "messageHeadline": "Switch some of the linux pipelines to use the new data download scrip\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 25, "deletions": 10}}, {"node": {"oid": "859a57d78166f2c7a020c82d7825e03b30e46e57", "messageHeadline": "Updated Dockerfile for OpenvinoEP (#1362)", "author": {"name": "avidiyal", "email": "41489828+avidiyal@users.noreply.github.com", "user": {"login": "avidiyal"}}, "additions": 142, "deletions": 40}}, {"node": {"oid": "93fb62bb3e891c8f631e10daa32eba979f87ded2", "messageHeadline": "More code cleanup (#1405)", "author": {"name": "Yuan Yu", "email": "yuanbyu@gmail.com", "user": {"login": "yuanbyu"}}, "additions": 9, "deletions": 54}}, {"node": {"oid": "a7b1a8969c08d5f812945aec08eaee1fe81857e1", "messageHeadline": "simply nocontribops-ci and fix build break (#1422)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 19, "deletions": 210}}, {"node": {"oid": "4383615cf68948829a1d273149926ab6be2d77bc", "messageHeadline": "implement conv+clip fusion (#1412)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 301, "deletions": 189}}, {"node": {"oid": "d2cc086bee2381cf6b35e718535523188c1f4930", "messageHeadline": "[OpenVINO EP] Minor bug fixes (#1388)", "author": {"name": "suryasidd", "email": "48925384+suryasidd@users.noreply.github.com", "user": {"login": "suryasidd"}}, "additions": 62, "deletions": 16}}, {"node": {"oid": "8720fe62e3a2a0b1c0d13eb249b2afe8a12b0223", "messageHeadline": "Added missing libraries to Windows wheel (#1415)", "author": {"name": "R. G. Esteves", "email": "rgesteves@gmail.com", "user": {"login": "rgesteve"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "c2aa2056b5af089a7be60f75e1f2c1ec107be0b2", "messageHeadline": "Sample for imagenet and batch prediction (#1372)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2810, "deletions": 12}}, {"node": {"oid": "d38badffdba37ad396af03849d245463ff302004", "messageHeadline": "Disable mklml in Windows Build", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a203077dcd7268fd3b80840b145592c18eda107a", "messageHeadline": "Relax timeout in CI system (#1394)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 6, "deletions": 0}}, {"node": {"oid": "07a2466d9f4cfde631811103f24d24ec511ef69f", "messageHeadline": "Use INFO instead of WARNING for an unused graph input. (#1235)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 19, "deletions": 3}}, {"node": {"oid": "fa4b956f12d0ad982a6bcb7201694a40cf8ffc20", "messageHeadline": "replace onnx:: with ONNX_NAMESPACE:: (#1376)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 5, "deletions": 7}}, {"node": {"oid": "61b733ce6d587a89b4d0dacdfdf6e6ac83ed1609", "messageHeadline": "Update optimizers to be able to utilize a constant initializer from a\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 253, "deletions": 162}}, {"node": {"oid": "d4ce31ea6d321c1cf21e386d22be5a0e2bcf97e7", "messageHeadline": "cleanup fused conv activation handling (#1403)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 305, "deletions": 353}}, {"node": {"oid": "c139e3ab33d14f9701ff4483f6a65a015adc2095", "messageHeadline": "Remove a few useless unique_ptrs (#1401)", "author": {"name": "Yuan Yu", "email": "yuanbyu@gmail.com", "user": {"login": "yuanbyu"}}, "additions": 52, "deletions": 74}}, {"node": {"oid": "719e58d831f91f9205119c5f22a7c68d1ec52539", "messageHeadline": "Use MLAS to retrieve the CPU preferred tensor buffer alignment (#1377)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 77, "deletions": 23}}, {"node": {"oid": "5a6f1c10d63f76d7c4b5d9cde1e26326e47a2f41", "messageHeadline": "Add OrtCreateStatus to the symbol list", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "3bf0e364e2efd865d2972dae7196b09e52e2f8cd", "messageHeadline": "Move CopyTensor out of IExecutionProvider interface. (#1268)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 459, "deletions": 253}}, {"node": {"oid": "e580b7630528dc551b91e6fa3a9c4af25f352457", "messageHeadline": "Fix ARM64 build + Add NuGet pipeline including ARM binaries (#1335)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 365, "deletions": 3}}, {"node": {"oid": "bfda9ca1c1e396411850e335f6c22ab1ffb90fe7", "messageHeadline": "Make sure submodule urls are up-to-date (#1357)", "author": {"name": "Maik Riechert", "email": "maik.riechert@microsoft.com", "user": {"login": "letmaik"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "20f6c84fd2ff3a6f68b62059860cef9cd48661bf", "messageHeadline": "Switch to use nvidia-docker2 command format", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "a7fcd60572febb7eea353605522899b1b85584bc", "messageHeadline": "Add missing 'openvino' option in perftest Usage message (#1367)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "aba7271ad7a8b2a5f2221eef9cc0d3e398f2145d", "messageHeadline": "Fix links (#1371)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "823fa3f39ca38ece197791726eb511d86ff43b56", "messageHeadline": "Integrate MLAS NCHWc support into ONNX Runtime (#1327)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 2132, "deletions": 61}}, {"node": {"oid": "42c18762f3825d0df9437ccf3f8345a605e962fa", "messageHeadline": "Update the log message for fallback case. (#1370)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "c483a1e3c6c21b72ebe6c46b21fe95b6898761d3", "messageHeadline": "Use simpler GEMM function for MatMul operator (#1365)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 48, "deletions": 256}}, {"node": {"oid": "57225cd4eeb44472b6556cfc6f45a407d479b90a", "messageHeadline": "Add C++ API test for NuGet package (#1364)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 26, "deletions": 2}}, {"node": {"oid": "298f30546b758fae102f0a28d7400336b382ac1e", "messageHeadline": "Fix the random UT failure for RNN/GRU cases which have padded sequenc\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "27da857b51c8c2f589052c378fb9fe48237d567b", "messageHeadline": "Fix an SAL annotation in onnxruntime_c_api.h", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6b32c77804579d3eb671029fd26400a537466f9b", "messageHeadline": "Dockerfiles for TensorRT, CUDA, build from source (#922)", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 198, "deletions": 14}}, {"node": {"oid": "3cae067a9b08521f7e04d3d0b383bfbe27746ddf", "messageHeadline": "fix non-standard u_int32_t type (#1358)", "author": {"name": "Maik Riechert", "email": "maik.riechert@microsoft.com", "user": {"login": "letmaik"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ac6a4afb0f6c5a8f443de1c73889ff3a9680b74a", "messageHeadline": "Add validation of shape when re-using a buffer in ExecutionFrame (#1356)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 58, "deletions": 2}}, {"node": {"oid": "58d6ff3f13aa2b7b36e3b9df0bdc766d3c53dd5d", "messageHeadline": "Remove AgentPool setting in CI yaml", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "3a588860cc4109b1b37e96eb8b299bc1739ba99f", "messageHeadline": "remove unused math routines (#1354)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 3, "deletions": 522}}, {"node": {"oid": "e9ce51ead46cdd43d56b889bed333e35bf6cb5ef", "messageHeadline": "Make GetTensorShapeFromTensorShapeProto return TensorShape and not it\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 20, "deletions": 15}}, {"node": {"oid": "5b93b02c69d8991c40260d6c4d593e5368891a0a", "messageHeadline": "Issue template update (#1339)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "b7ae0d56947bd981c7abcdc18ff69b9f01bc2480", "messageHeadline": "Fix link (#1351)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "93528d9b3cdd08639d10cfe8398b607b9efc40ba", "messageHeadline": "Reduce memory footprint of nGraph (#1296)", "author": {"name": "R. G. Esteves", "email": "rgesteves@gmail.com", "user": {"login": "rgesteve"}}, "additions": 126, "deletions": 0}}, {"node": {"oid": "9f9ff19bdc558c31581ebf6493eb7e1c6eb1b4f2", "messageHeadline": "Copy shared library after build ORT Server (#1347)", "author": {"name": "NonStatic", "email": "nonstatic@engineer.com", "user": {"login": "non-static"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "2714576d0ae89f553874394d12c1f3cb4724a6de", "messageHeadline": "Update ONNX Runtime server doc to reference Jupyter notebook (#1340)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "a8ff209ab6587d6276d7b30b9dd861d27320b3af", "messageHeadline": "Refactor Onnx runtime Server to only use public APIs (#1271)", "author": {"name": "Colin Versteeg", "email": "cverstee@gmail.com", "user": {"login": "csteegz"}}, "additions": 1699, "deletions": 603}}, {"node": {"oid": "e3919d3fcecbdd80f017f8176d4d5435d05a4d38", "messageHeadline": "Cleanup naming of test input to use .onnx for models. (#1337)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 45, "deletions": 45}}, {"node": {"oid": "0d204f3f067b389e97c7f7576d38e4f54edac913", "messageHeadline": "Implementation of TVM codegen library (#888)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 6788, "deletions": 374}}, {"node": {"oid": "9d3b6b3a49aad727624f49efa728b3b7625082c8", "messageHeadline": "Disallow overriding initializers if IR version < 4  (#1324)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 305, "deletions": 122}}, {"node": {"oid": "2a6c69de2b2f1fd94fd5162f08812d9ee8be2fda", "messageHeadline": "Implement the Concat CUDA kernel (#1333)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 204, "deletions": 51}}, {"node": {"oid": "5e54bbffecb3a7cbc5e3d354fbef411698c00e08", "messageHeadline": "PyOp documentation Revisions (#1318)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 26, "deletions": 22}}, {"node": {"oid": "1bf80e30fa2f1b93c98849fbfa336be9f6ef5325", "messageHeadline": "Ryanunderhill/MNIST sample (#1330)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 295, "deletions": 0}}, {"node": {"oid": "bf6a9f9c27e14cf334e6cf30e8a91c39f2456728", "messageHeadline": "Rashuai/py op example (#1325)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 45, "deletions": 1}}, {"node": {"oid": "c65489a47fde37330033a6b257d614bb78304d73", "messageHeadline": "Initial PR for NNAPI execution provider (#1220)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 618, "deletions": 1}}, {"node": {"oid": "98ea675e40ce8283d550674c1e9d33e4e0b7c054", "messageHeadline": "Fix typo: op[s]iops -> op[t]ions. (#1329)", "author": {"name": "xkszltl", "email": "xkszltl@users.noreply.github.com", "user": {"login": "xkszltl"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f67a1629fcb6ce24ad54fb33b19709c2b30e9c66", "messageHeadline": "Documentation reorganization (#1143)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 140, "deletions": 74}}, {"node": {"oid": "5af2aec3a23350472cb385cb8ade0955378e8468", "messageHeadline": "[OpenVINO-EP] use logging infrastructure to display EP log messages (\u2026", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 79, "deletions": 55}}, {"node": {"oid": "28759e2f6ffc5c33f20cc17d1e68e5f532fe5130", "messageHeadline": "Uninstall the preinstalled cmake in tensorrt image because it's too o\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "a077ac8df572ab5cf3488da5cd5d1183cdb23b4c", "messageHeadline": "Support non-default negative axis value and intuitive data type combi\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 53, "deletions": 7}}, {"node": {"oid": "2698edbc98950c31edbae91af8ca049f61c46eb6", "messageHeadline": "enable tests (#1310)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "2f698bd54b713bb87dbd0bbb913e94bcf7fd480c", "messageHeadline": "Fix NMS const_cast that modified kernel state creating (#1303)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 249, "deletions": 145}}, {"node": {"oid": "04d581995ddaf0af0cb9744ebbc62020a9c5e573", "messageHeadline": "Use manylinux2010 image to build linux python wheels (#1282)", "author": {"name": "Matthieu Darbois", "email": "mayeut@users.noreply.github.com", "user": {"login": "mayeut"}}, "additions": 327, "deletions": 145}}, {"node": {"oid": "0951f53c801846735fed861081f19b37f9871e15", "messageHeadline": "Update ONNX to d94f99d21a9a0820d58966410ceaf525132f85f1 to pickup cha\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 8, "deletions": 6}}, {"node": {"oid": "59de37af1fb74aaa8f78c5a376dfcdb8fe27e064", "messageHeadline": "Add CUDA Expand operator (#1292)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 237, "deletions": 2}}, {"node": {"oid": "a79ab5ec5b06874ce581875a2ba1835c48458d39", "messageHeadline": "Add document for ONNX Runtime latency profiling and JSON file viewing\u2026", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 14, "deletions": 0}}, {"node": {"oid": "b8d370029f38649e5c35b58b9970e94003e7cd29", "messageHeadline": "Check that specific inputs are constants in Conv(Add|Mul|BN)Fusion ru\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 114, "deletions": 23}}, {"node": {"oid": "089b1ef3bbb465ba286f1197d9fbd041aab0c09b", "messageHeadline": "Enable max/average pooling onnx_test_runner tests (#1129)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "05a222a961c4729c729c24cbbbc19a12dde6e211", "messageHeadline": "enable quantization tests (#1293)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "3ebad81abc831d6e1b72548a71da7d59cca651b5", "messageHeadline": "MLAS: NCHWc low-level changes (#1283)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 10988, "deletions": 906}}, {"node": {"oid": "a462328d9db70afac3eafffa4038d871b4aa1847", "messageHeadline": "Handle case where the Loop 'M' and 'cond' inputs can be considered sc\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 50, "deletions": 15}}, {"node": {"oid": "c0cf2213bc57699e3c6357420d3c6a15ad78d6b7", "messageHeadline": "Parallelize TreeEnsembleClassifier batch predition (#1276)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 11, "deletions": 9}}, {"node": {"oid": "a56b294428699f179a9ae25083d87173382de027", "messageHeadline": "Activate compliance tasks for private builds, and also set a daily sc\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 35, "deletions": 1}}, {"node": {"oid": "a542769b50af1df42a0f903e3a9adc25ac230538", "messageHeadline": "Fix misleading notion that Flatten op is not supported for opset 10 (\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 27, "deletions": 5}}, {"node": {"oid": "86dc3b4360718aaca90cf76f0c40187bb2c016a0", "messageHeadline": "Fix bug in the transformer that removes unnecessary Cast nodes where \u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 68, "deletions": 20}}, {"node": {"oid": "c9d83a52a8dc1c3c2a09d61f5f6435b92125aa68", "messageHeadline": "Implement contrib op CropAndResize (#1277)", "author": {"name": "David Fan", "email": "30608893+jiafatom@users.noreply.github.com", "user": {"login": "jiafatom"}}, "additions": 620, "deletions": 35}}, {"node": {"oid": "06642dbbacd8f7af20a98ac5ca4ce418e69bce41", "messageHeadline": "selectively enable quantization tests for ngraph (#1290)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a571ea74a6ffe9e289c57079502415eda9987090", "messageHeadline": "update onnx (#1287)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "01715c0ff18d4d845dc15c99e24a9d7e6513cd8e", "messageHeadline": "update doc \"How_To_Update_ONNX_Dev_Notes\" (#1288)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 12, "deletions": 13}}, {"node": {"oid": "92dc5c506dbf2cc35df598d018070c29c90511a3", "messageHeadline": "move all contrib ops to contrib ops namespace (#1190)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 955, "deletions": 646}}, {"node": {"oid": "12d70abd2954fe9eded04726df1c9474e4506b7d", "messageHeadline": "Propagate errors from parallel executor (#1262)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 194, "deletions": 30}}, {"node": {"oid": "671c15a56aaab6daf327c5b094c11a3547b16ab9", "messageHeadline": "Treat attribute warning as non-error on cross compiling ARM (#1261)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 5, "deletions": 2}}, {"node": {"oid": "2bfbcd323a69ff101b6f42013838051edddc957e", "messageHeadline": "fix OrtValue Release condition (#1182)", "author": {"name": "Yufeng Li", "email": "roli@microsoft.com", "user": {"login": "yufenglee"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "204bd38d6a4d40d13affd234ec55459e354f1f4a", "messageHeadline": "Add ability to set graph optimization level in onnx_test_runner. (#1275)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 28, "deletions": 9}}, {"node": {"oid": "c8db2d507eafe8da73ebd22e8a07cd4494261141", "messageHeadline": "Actualy add the C++ headers to the nuget packages (#1267)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4d765dc6d00a631cf60f76c271a1ae285ed57e0a", "messageHeadline": "Return error message from status instead of swallowing it. (#1221)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 34, "deletions": 20}}, {"node": {"oid": "18b7d2b18a59b913ff37273aba24a75b4f323981", "messageHeadline": "Add document of ONNXRuntime performance tuning (#1266)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 94, "deletions": 0}}, {"node": {"oid": "3b71701f9164ca84095efab9bb5b009d2a8b50fd", "messageHeadline": "Update onnxruntime server docker file with ONNXRUNTIME_VERSION in cma\u2026", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "3275e44c62c3c7221394935b2766836e1b6711b2", "messageHeadline": "Change the memory alignment for default cpu provider (#1269)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "eb833057d24d7f7c87a239352e2c2e3bdbac9283", "messageHeadline": "Update Model_Test.md (#1264)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "d3e5474c1d4624aefb9f06c710d32fe486e3a0af", "messageHeadline": "Refactor CI pipelines - add GPU NuGet pipelines and ESRP code signing\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 425, "deletions": 307}}, {"node": {"oid": "766c6b61633c3b66e523d0b0184ed57aca1768bc", "messageHeadline": "Add an API for retrieve ORT version (#1263)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 4}}, {"node": {"oid": "5b87f07d8093c2b28eb20684ed2594e5af9838a2", "messageHeadline": "enable matmul tests (#1255)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 1, "deletions": 6}}, {"node": {"oid": "ba25ea3643153370be4300da725dc3057d9128c7", "messageHeadline": "Allow building Docker container based on a different git repo. (#1222)", "author": {"name": "Jorgen Thelin", "email": "jthelin@microsoft.com", "user": {"login": "jthelin"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "df68111b9809d9ec2db58633ebba5fcf0ff8f47b", "messageHeadline": "Fix a bug that fused func manager in subgraph session state is nullpt\u2026", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 27, "deletions": 17}}, {"node": {"oid": "23838d9c2a637c5eacafe29194c73eefd2f4943f", "messageHeadline": "Add enable/disable mem pattern api for python and csharp. (#1227)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 20, "deletions": 0}}, {"node": {"oid": "e26e11b9f7f7b1d153d9ce2ac160cffb241e4ded", "messageHeadline": "Quantization tool to support quantization of Conv and MatMul nodes. (\u2026", "author": {"name": "PhaniShekhar", "email": "30535191+PhaniShekhar@users.noreply.github.com", "user": {"login": "PhaniShekhar"}}, "additions": 1159, "deletions": 0}}, {"node": {"oid": "051ee681a3ed6282c56fdd4ab82b702edb2a24c2", "messageHeadline": "Add a few suggestions to coding guideline (#1238)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "c96049fe4ad16f1964257e442fad6906e0acb198", "messageHeadline": "Update ONNX version to include new fixes/changes (#1250)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 14, "deletions": 4}}, {"node": {"oid": "6477d4e7564f0d8b6203544abfc348b18c3a0b98", "messageHeadline": "Return better output shape for Loop with zero iterations (#1233)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 61, "deletions": 21}}, {"node": {"oid": "a4148c85a567ad74a4ed9ef5f3772f47693153bf", "messageHeadline": "call install_onnx.sh with relative path (#1225)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2d0e5231da465ff46b08e1429072de8bc7043ac4", "messageHeadline": "Fix the build issue with gcc 9.x (#1242)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 2}}, {"node": {"oid": "08ff75930b8ef9b8e1fc6a6e47a1c0cadd3e0778", "messageHeadline": "Add pull request template (#1249)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "aae24f7f9e78a7f67fcf4992722b0f79c0fb1364", "messageHeadline": "Return right category of error. Use INVALID_ARGUMENT when the input i\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 42, "deletions": 44}}, {"node": {"oid": "48df19f53d155592ce52a4fd58dbac5f4af53632", "messageHeadline": "perftest: verbose logging if -v is given", "author": {"name": "Maik Riechert", "email": "maik.riechert@microsoft.com", "user": {"login": "letmaik"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "8d15ffd8f5d06cd1b3cb87181bf7ba05e50f0560", "messageHeadline": "Initial commit for OpenVINO Execution Provider (#935)", "author": {"name": "S. Manohar Karlapalem", "email": "manohar.karlapalem@intel.com", "user": {"login": "smkarlap"}}, "additions": 2748, "deletions": 27}}, {"node": {"oid": "f22d1df04f582402431ba4e65127af16688fead1", "messageHeadline": "Fix a bug that constant folding doesn't check the returned value of O\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "723e30b361101b24efae02855d479213ac24db47", "messageHeadline": "Refine node selection logic in ShapeToInitializer optimizer (#1219)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 39, "deletions": 18}}, {"node": {"oid": "e84cb7b579d250c86223d70def7505a65d464891", "messageHeadline": "Revert \"Use shared threadpool in LSTM (#1167)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 11}}, {"node": {"oid": "40ec69191acf4d7761e087176b973123cb8c9b37", "messageHeadline": "Revert \"Let mlas use the session threadpool for gemm functions (#1196)\"", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 234, "deletions": 231}}, {"node": {"oid": "13d855840406cde2a2e9406a0f53a75bbd96be77", "messageHeadline": "move environment.h/cc from framework to session project/folder. (#1241)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "d448f39832d1d881bf781c8d870e08499468209b", "messageHeadline": "refactor a little bit to not ask every execution provider to give an \u2026", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 6, "deletions": 16}}, {"node": {"oid": "6d7081f33f8f4f414c12c8279979bf00586c9ec0", "messageHeadline": "Add PyOp doc and UT (#1200)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 143, "deletions": 32}}, {"node": {"oid": "f9374217c6e866c8027458639e3d50e4c20efb29", "messageHeadline": "Add -P flag to perftest & repurpose -x (#1236)", "author": {"name": "Maik Riechert", "email": "maik.riechert@microsoft.com", "user": {"login": "letmaik"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "5f21eedcbdf9bd329512ada83b0c0266f624f286", "messageHeadline": "Script for uploading code coverage data to dashboard (#1209)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 165, "deletions": 5}}, {"node": {"oid": "08731589c901d57180b140748107bebb73486a85", "messageHeadline": "Refactor CI pipelines, and add YAML NuGet package generation pipeline\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 1122, "deletions": 328}}, {"node": {"oid": "2148827d3a82916da45a24a99a52996ed56772c5", "messageHeadline": "Implement CUDA kernel for Shrink operator (#1208)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 208, "deletions": 1}}, {"node": {"oid": "6850e5596629852a6789ff7aa8e15aeb813ed76a", "messageHeadline": "Fix install_onnx.sh issue (#1201)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 18, "deletions": 17}}, {"node": {"oid": "fc9a895b46fa1730dedf6dcbb13e8b6ede20960e", "messageHeadline": "Add shape inference logic for Crop (contrib) op (#1157)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 140, "deletions": 26}}, {"node": {"oid": "6c17567d7bfabacae2991175358ddc4b19375445", "messageHeadline": "Add C++ headers to nuget package (#1218)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "065e9dc1baceb5fc27af8babadafec51e2684a77", "messageHeadline": "Block size mismatches are expected if sequence length varies or there\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "fa2eea73399c9625fa79761a4b84f240e0ec9669", "messageHeadline": "Use cmath instead of math.h for fabs (#1217)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "c1a34a8ba60811838927e054504669fca1fb2038", "messageHeadline": "Add ability to dump node input/output (#1202)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 150, "deletions": 2}}, {"node": {"oid": "bbdd1d658b7e83a5c56dcd80980312125f37e731", "messageHeadline": "use cudnnRNNForwardInferenceEx for unpacked (padded) layout case (#899)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 212, "deletions": 118}}, {"node": {"oid": "38963d81eb405b47a9b71ba4247ae87a74f70550", "messageHeadline": "Simplify some CustomOp code (#1206)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 16, "deletions": 39}}, {"node": {"oid": "87d65389e6eb197fd71de24f5e30b5b15ca3780d", "messageHeadline": "Add ability to change the logging severity of the default logger. (#1\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 158, "deletions": 93}}, {"node": {"oid": "15bcde5053d5ab482c49bad6de2eb832ab83fe54", "messageHeadline": "Fix a build break in tf_test_session.h (#1205)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "6d5ea08936b95558e4601a453442c970c9d69f96", "messageHeadline": "Add nsync lib to onnxruntime_mlas_test's deps (#1199)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 3}}, {"node": {"oid": "0b9b429fe1fd95e1738976ddbced9246890b2abc", "messageHeadline": "CPU GRU and LSTM Ops: Address corner case where output is uninitializ\u2026", "author": {"name": "edgchen1", "email": "edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 94, "deletions": 4}}, {"node": {"oid": "24d6b0f5c48a724f2d985dd3f96cf8c026e95934", "messageHeadline": "MKL-DNN Subgraphs (#1116)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 3239, "deletions": 2}}, {"node": {"oid": "3c3186c761463b9320b94bd02f37b303d0069494", "messageHeadline": "Convert more C APIs to return OrtStatus (#1194)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 324, "deletions": 214}}, {"node": {"oid": "011957995ee23c3a7514cfc91769e8c7a974c696", "messageHeadline": "add immutable exclude list (#1181)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 59, "deletions": 39}}, {"node": {"oid": "a92998c235ee280013ac6ee0fd304a7458b4d0c8", "messageHeadline": "Uncomment ConstantOfShape tests. (#1059)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 6, "deletions": 34}}, {"node": {"oid": "f72c1501c5dc48e45e8b56bb6ccec7724993f08c", "messageHeadline": "Optimize concat PrepareForCompute perf by reserve vector size. (#1198)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "97dfd5ee2167987eb1d9cb98488195ed44b004a2", "messageHeadline": "Add code coverage (#1192)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 134, "deletions": 8}}, {"node": {"oid": "280ab9a2d09e626f4bf44531758b2b290cab0cd2", "messageHeadline": "Let mlas use the session threadpool for gemm functions (#1196)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 231, "deletions": 234}}, {"node": {"oid": "be36385a8c5796ecedc7a59f1760b5004e69e819", "messageHeadline": "Delete docker/scripts/install_deps_x86.sh and enable onnx tests for x\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 72, "deletions": 187}}, {"node": {"oid": "6b586bc041cfac1e7e36cf416e5f25d7a5daeed5", "messageHeadline": "Avoid warning status in python release pipeline (#1195)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 0, "deletions": 19}}, {"node": {"oid": "71466fc8051365e38ec901ab680968a4374454eb", "messageHeadline": "Fixes two broken links. (#1081)", "author": {"name": "gapar2", "email": "571694+gapar2@users.noreply.github.com", "user": null}, "additions": 0, "deletions": 0}}, {"node": {"oid": "ccab8165ebfca437c4640ef2d72a3279c4b10324", "messageHeadline": "Delete scripts/install_ubuntu_x86.sh (#1189)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 26, "deletions": 82}}, {"node": {"oid": "32c6c71e8678df299f06fbeeef26f46646a830bf", "messageHeadline": "Convert Shape operator to initializer (#1159)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 219, "deletions": 24}}, {"node": {"oid": "cdb27de09098c3048894c590c612d80282250266", "messageHeadline": "implement python opeartor (#1045)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 720, "deletions": 2}}, {"node": {"oid": "8d68098c208418d47be8a4935fd1c3eb01faa205", "messageHeadline": "ConstantOfShape CUDA implementation (#1168)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 218, "deletions": 134}}, {"node": {"oid": "e43e64bf844eb9719367a3cd9d97a0bb357a2f7f", "messageHeadline": "Implement Equal for CUDA. (#1183)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 64, "deletions": 0}}, {"node": {"oid": "d33dbb23b23c1d1f5f9ab76091ad412ed1d055eb", "messageHeadline": "replace onnxmltools by keras-onnx in one example (#1151)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d8ac0d64d07898dcea8ea7ed84f4602ec404731d", "messageHeadline": "Make C API capable of defining CUDA custom ops (#1178)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 42, "deletions": 2}}, {"node": {"oid": "b68bb51dd0e3927fe62c83e2d20eb08e1f0e16f5", "messageHeadline": "Change SessionOptions APIs to always return a status (#1171)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 112, "deletions": 95}}, {"node": {"oid": "b23ab6a06e6d1a07c7304e9f7d620741f55a3b82", "messageHeadline": "Implementation of sparse tensor (#1121)", "author": {"name": "G. Ramalingam", "email": "grama@microsoft.com", "user": {"login": "gramalingam"}}, "additions": 910, "deletions": 17}}, {"node": {"oid": "7a80770b524d247c0418ed2edb21c700a55d4f46", "messageHeadline": " Remove filtering from the backend scripts runner (#1169)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 23}}, {"node": {"oid": "1a86421affa7a3cd677931a5b8e88cf81e9ee086", "messageHeadline": "Create a syslog sink for logging in !Win32 env (#1163)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 102, "deletions": 7}}, {"node": {"oid": "88ea58a38376aff58c0adf1ec0dd221ac42ec9ac", "messageHeadline": "Add double tensor support for Div and Sub (#1172)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 0}}, {"node": {"oid": "c6abb17b8d6484dafb5d05c0cdbf8a725cea286b", "messageHeadline": "Use shared threadpool in LSTM (#1167)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 11, "deletions": 10}}, {"node": {"oid": "7cd2d9f3c41ede8427fb5c4270eceae678e8638a", "messageHeadline": "Change Path_lib to not use Shlwapi.dll if compiled as WINAPI_PARTITIO\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 29, "deletions": 1}}, {"node": {"oid": "7c4494a0bc3baad95a783c6759c753f6b5a51b5e", "messageHeadline": "Fix CUDA thread_local to allow multiple CUDA execution providers (#1149)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 53, "deletions": 61}}, {"node": {"oid": "a863c67ef873c3ac2f3055d1dfb237535d9c1cb2", "messageHeadline": "Fix error message for legacy opset (6 and lower) models at model load\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 18, "deletions": 3}}, {"node": {"oid": "7316e54153e8189da7922c00004f0fba427088dc", "messageHeadline": "re-enable disabled tests on nGraph after fixing remaining subgraph re\u2026", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 16, "deletions": 12}}, {"node": {"oid": "148141dd5fc0651da6ed8f7522d8ee57a48cde5f", "messageHeadline": "Change Compute function to return a status code instead of an integer\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 22, "deletions": 28}}, {"node": {"oid": "c18de6817bb6afd02c5fe6bf3afc5330e98a978e", "messageHeadline": "Rename MLValue to OrtValue (#1154)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 202, "deletions": 216}}, {"node": {"oid": "b8a699f70b8704093d41918643becfd306d576d6", "messageHeadline": "Update MaxPool & AveragePool to support opset 10 (#1141)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 58, "deletions": 20}}, {"node": {"oid": "10ea77a3d10a486a6755f77a17af44b36a880ff9", "messageHeadline": "add details aboud adding execution providers in the C api to comments\u2026", "author": {"name": "Torkel", "email": "torkel@voysys.se", "user": {"login": "torkeldanielsson"}}, "additions": 8, "deletions": 4}}, {"node": {"oid": "6c9d815de592a77d83f5a5e2d66ce12c43bfa4da", "messageHeadline": "Revert \"Remove openmp flag (#1140)\" (#1146)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 62, "deletions": 33}}, {"node": {"oid": "71cbd7679cad585681f29881684a37808e2e6985", "messageHeadline": "implement cuda slice opset 10 (#1137)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 41, "deletions": 11}}, {"node": {"oid": "a7137a0f9d1795fa86a25a726a3c638d17085b93", "messageHeadline": "Remove openmp flag (#1140)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 33, "deletions": 62}}, {"node": {"oid": "05110a65586a86f5b671a850b93dc0434589af43", "messageHeadline": "Adding custom op ConvTransposeWithDynamicPads. (#638)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 248, "deletions": 31}}, {"node": {"oid": "6c408c3a75607ea622565312dbae1a160a1f6949", "messageHeadline": "Simplify ONNX Runtime Server CI build (#1136)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 1, "deletions": 40}}, {"node": {"oid": "facdf77f84a79ef38e8732cf0d898b8ee6d61897", "messageHeadline": "Fix shape inference bug in GatherND contrib op (#1132)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4757933afe56ea4332141ebd8e2feed204cd915f", "messageHeadline": " Exclude test by onnx version tag (#1073)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 168, "deletions": 117}}, {"node": {"oid": "f9f6818e4cd6e2badff68f4d55e34b50f6e151fb", "messageHeadline": "Add comments and organize the C++ header into the main header plus a \u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 469, "deletions": 426}}, {"node": {"oid": "66a6f2b0e39a8eb5006440778323abe6fb90686e", "messageHeadline": "Add a check for onnxruntime_USE_MKLML AND onnxruntime_USE_OPENMP (#1131)", "author": {"name": "Changming Sun", "email": "me@sunchangming.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "4e231ad90792156b5188ba57954ef0aa2e19013d", "messageHeadline": "Split binary/symbol and then upload ortsrv nightly build to blob stor\u2026", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 54, "deletions": 5}}, {"node": {"oid": "2072d34d156ab5fbfff82ee4481733ab19951670", "messageHeadline": "Fixing broken links for graph transformations in High Level Design do\u2026", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "f51b081ec2db36513ba2d84dee26472681d2d1b0", "messageHeadline": "Revert \"fixing clang build failure\" (#1055)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ef663950606b206c3a09d7559556b595ac5477df", "messageHeadline": "remove unnecessary lock and code clean up (#1114)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 13, "deletions": 48}}, {"node": {"oid": "eda4c5cb47379e053e9a49caec5f1deba1181ad7", "messageHeadline": "fix ngraph build (#1124)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "8c7e4eb3fbf7ba4093a1d2f30e27bfc171bbc957", "messageHeadline": "Fix run_model api. (#1111)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 29, "deletions": 11}}, {"node": {"oid": "e19bc2d07442ccf9a1f3fb654b3c30353a39c617", "messageHeadline": "Raise max request size in ORT server (#1119)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1ea3e8633ca118b3e97d7ba9e68e8a4b5531fe28", "messageHeadline": "CUDA opset9: Update Cast/MatMul version, add Erf (#1106)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 213, "deletions": 77}}, {"node": {"oid": "71560843f86a424c24a1900f756ac2de37f52467", "messageHeadline": "Add Boost dependency for onnxruntime server. (#1112)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 107, "deletions": 70}}, {"node": {"oid": "2cf56639edbfbb4a89b738ae0798bb34e6af5992", "messageHeadline": "Minor update to NuGet package tests -- allow model download in separa\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 21, "deletions": 20}}, {"node": {"oid": "b9e8aac5f6760acf849dec4c0bd13554bff45ce8", "messageHeadline": "Enable some more tests (#1101)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 64, "deletions": 54}}, {"node": {"oid": "8808efd9e34ae0692255e0d07e37bd2acfb545b8", "messageHeadline": "Add zero size check before setting thread pool so that the zero defau\u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "f6df36b68b59b88ac2f6f6fa98a171fde28b8793", "messageHeadline": "Add rewrite rule to handle Relu + Clip (#1105)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 168, "deletions": 15}}, {"node": {"oid": "b54a292ba265e65b6a4499f8d6f49010db932f47", "messageHeadline": "Add version and latest commit id to ONNX Runtime Server (#1078)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 41, "deletions": 1}}, {"node": {"oid": "ea29a664cd974b407752ff5251d436ddc040020c", "messageHeadline": "Fix android build when API<23, fix android test, update build doc and\u2026", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 21, "deletions": 6}}, {"node": {"oid": "2d92b95729c530626e52ff8f3df93e814f5e6ef4", "messageHeadline": "Enable Unsqueeze elimination (#1104)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "ee6217972bd8c5a391dbdaf7b0349699c4efa392", "messageHeadline": "Fix when rewrite rule gets registered to multiple op types; update co\u2026", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 54, "deletions": 48}}, {"node": {"oid": "9129a652c598307e7e959c0925b6b839c5a0529c", "messageHeadline": "Ryanunderhill/cxx api2 (#1091)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 295, "deletions": 367}}, {"node": {"oid": "723d5c782a5826bcb101052e9c4b9a623cfe0712", "messageHeadline": "Improve TensorRT GetCapability to Enable More Models (#1012)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 462, "deletions": 291}}, {"node": {"oid": "b44a30bca75a4d30cca957e348b48d316ac50cf4", "messageHeadline": "CUDA CPU/GPU sync optimization (#1100)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 56, "deletions": 13}}, {"node": {"oid": "cd9b9e6102757277e9761421ce9efa04753cd78e", "messageHeadline": "Accelerate CUDA transpose (#1102)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 59, "deletions": 4}}, {"node": {"oid": "f4a9ccae999f8478c805a74318a699ea87238955", "messageHeadline": "Enable nGraph Debug ci test. (#1000)", "author": {"name": "R. G. Esteves", "email": "rgesteves@gmail.com", "user": {"login": "rgesteve"}}, "additions": 76, "deletions": 9}}, {"node": {"oid": "4a8d75386b9fdfac73724b50af06520a52bf3e2e", "messageHeadline": "Clarify/state expected usage of non-const references and 'auto' in co\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 7, "deletions": 1}}, {"node": {"oid": "11243253f234e6b46d894acf8b112b6530edeafe", "messageHeadline": "Disable mvn model test when contrib ops are disabled since this model\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "c7d1c007d592d33194cc63244f91a4446147b4ae", "messageHeadline": "Fix accidental copy where a reference was fine. (#1090)", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "0157b6c209db1ae0e516e618a1ba89b78b1ce6e7", "messageHeadline": "Enable specifying tests for onnx_backend_test_series.py to run via th\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 91, "deletions": 63}}, {"node": {"oid": "7421755198cefc4680332d52db3e444fa9d463f8", "messageHeadline": "Optimize ExecutionFrame to avoid mem re-allocation. (#1085)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "055f5af78de40285dc46fe6256291697800b36cb", "messageHeadline": "Remove path change in build.py (#1089)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "d865d3496863d76650fcdfa196c72e427b252ec3", "messageHeadline": "Fix CUDA crash in fast RCNN model (#1092)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 30, "deletions": 9}}, {"node": {"oid": "2a07dc2fbfbb9b83272256020f922779c7f783e2", "messageHeadline": "Add provider information for kernel in perf test. (#1082)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a6a4fef677be4369e8e329953810150f38b180cd", "messageHeadline": "Support both Scalar tensor and 1-d tensor for Loop iter/cond input (#\u2026", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 92, "deletions": 41}}, {"node": {"oid": "376a8240ac7005357f64a6994de05b98e4b842f9", "messageHeadline": "Add x86 legs to CI builds (#1005)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 599, "deletions": 0}}, {"node": {"oid": "602028ef19ed3af69ea4e9f4900dfd4cc898f5e0", "messageHeadline": "Remove tests from exclusion list  (#1077)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 0, "deletions": 6}}, {"node": {"oid": "bc01f7b16f9e0bb490b556c5cf4b29e596a9d175", "messageHeadline": "Ensure the Eigen code that is getting #included is licensed under the\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "9f79ff52ba2241a6f808a38a27346f559184e14c", "messageHeadline": "update nuget version (#1075)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9673f3d4940ac5eca9425c300c3e16e2836a04c4", "messageHeadline": "Jignparm/minor update linux test (#1074)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "c7cb0c052dd1ca24781eea609930ea0f0ce48712", "messageHeadline": "Add the onnx inference on AKS (Azure ML) notebook from //build (#1071)", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 877, "deletions": 0}}, {"node": {"oid": "c124f7e81c6c2219c6fd4f1b25917574ca6b596a", "messageHeadline": "Uncomment mvn test. (#1072)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "2e01018e332f615369a93cf6953c1015e294064d", "messageHeadline": "Fixes #1007, base_value in TreeEnsembleClassifier (#1015)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 202, "deletions": 70}}, {"node": {"oid": "c1b2bd937cccd54188fdb733b9b35545c2398e26", "messageHeadline": "Add -M and -A arg to onnxruntime_perf_test (#1070)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 23, "deletions": 7}}, {"node": {"oid": "bf6f19c6b729ee2b5e067f48b30bb412951b828a", "messageHeadline": "Use auto when possible (#1058)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 226, "deletions": 230}}, {"node": {"oid": "15561581f9d65c787a458a37ba5af7116e9cc0ed", "messageHeadline": "Change the INCLUDE path for gemmlowp (#1066)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "64aa55eac8908cd359fb314e648443144734eb8c", "messageHeadline": "Re-order InferenceSession::session_state_ in the member declaration (\u2026", "author": {"name": "Tao Qin", "email": "48697806+TaoQinMS@users.noreply.github.com", "user": {"login": "TaoQinMS"}}, "additions": 9, "deletions": 7}}, {"node": {"oid": "a42222f9deef0b0ef810f9bc8469c7c30d3dba31", "messageHeadline": "bump onnx version & fix conv/pool tests (#1067)", "author": {"name": "Bowen Bao", "email": "bowbao@microsoft.com", "user": {"login": "BowenBao"}}, "additions": 11, "deletions": 7}}, {"node": {"oid": "3209204ef2ccd418831aa5f00cee1283bf8a90f6", "messageHeadline": "Uncomment tests. (#1069)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "2ad626752acb8226097b27fefdb1e227c01c3bd6", "messageHeadline": "Use local ort python package in server model tests (#1027)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 31, "deletions": 17}}, {"node": {"oid": "32da12491dc55ae41bba65197b64eecf0f36d222", "messageHeadline": "x86 support for C# API (#962)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 224, "deletions": 129}}, {"node": {"oid": "3a32b0eb9966db6190014046fc114275f184572c", "messageHeadline": "Change function kernels to use CustomOp APIs (#1020)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 155, "deletions": 230}}, {"node": {"oid": "7763f62a0970b6713ae8e6908490a1f899161545", "messageHeadline": "fix ngraph build (#1062)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "58d9f6a1458cfdedea29822544185982514aa713", "messageHeadline": "Enable greater/less tests. (#1061)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1, "deletions": 5}}, {"node": {"oid": "f90bd234ff86a872ab2796e1821ee0b003c089e8", "messageHeadline": "fix warning in scatter op (#1054)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fd5eb3b351780e34893c7c3ccf2319db1846e58d", "messageHeadline": "Remove dropout nodes during inference (#1018)", "author": {"name": "Doris Xin", "email": "doris.s.xin@gmail.com", "user": {"login": "dorx"}}, "additions": 183, "deletions": 5}}, {"node": {"oid": "2663b9c44381b30525ae6234e13ed25c69206d07", "messageHeadline": " Remove unnecessary casts from OrtValue to MLValue(#1051)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 965, "deletions": 1152}}, {"node": {"oid": "a7039601c4bf469131569d0c1261f13cd6bd67d9", "messageHeadline": "setting input/output for partitioned function subgraph (#1019)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 108, "deletions": 124}}, {"node": {"oid": "11069765dc0dae312304bb7545c5ed1765bed72a", "messageHeadline": "Fix C-API sample (causing internal build failure (#1047)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "99556b111d2d5f7a722d21b504c9d8a0122a18a6", "messageHeadline": "Make MemPatternPlanner on/off switchable in model weight loading  (#989)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1562, "deletions": 1780}}, {"node": {"oid": "7bce3771135ec738cd9db28beaa060c62bb53f6d", "messageHeadline": "Fix LTO build failure on ubuntu (#1048)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 19, "deletions": 14}}, {"node": {"oid": "2324f65a22c27858599a6d22d1437ac273f6d51d", "messageHeadline": "Fixed typo on paragraph 14 (#997)", "author": {"name": "Dar\u00edo Here\u00f1\u00fa", "email": "magallania@gmail.com", "user": {"login": "kant"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e4225f8234272d039f22c2094d83a7fc6df7ad61", "messageHeadline": "Combine OrtValue and MLValue into one type (#1043)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 74, "deletions": 80}}, {"node": {"oid": "6c65640708c6a674ce3f0a0bee6348601ca9ad3f", "messageHeadline": "Update SessionState::GetKernel to just call find() instead of count()\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "2e19b14e4e8b146f91e7ab63968866e092bbdf01", "messageHeadline": "Fuse elementwise add/multiply into convolution (#1028)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 56, "deletions": 15}}, {"node": {"oid": "d14e65a224d9fc2867856f222c190080f75379d4", "messageHeadline": "Finer control over when Python tests are run (#1023)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "303f3c1278157d32cbb0b3a14b6cc9cee2e56038", "messageHeadline": "drop GIL before Run to enable multiple python threads to execute in p\u2026", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 33, "deletions": 32}}, {"node": {"oid": "d3c0cd24fcc97f7fe25950fd8508519b21bbf67f", "messageHeadline": "Change the cpu allocator info used in OptimizerExecutionFrame class (\u2026", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "5062be612e4466240ab9f07a979a71975de30691", "messageHeadline": "update (#1041)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 25, "deletions": 5}}, {"node": {"oid": "fc17a8bc08607270828b7a3e5f134c8e895e36d9", "messageHeadline": "Fix MIN and MAX aggregation functions (#1022)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 90, "deletions": 15}}, {"node": {"oid": "875c4c2f3fb664b8e846f051444a92a7e45f73ea", "messageHeadline": "restore ninja compatibility", "author": {"name": "Maik Riechert", "email": "maik.riechert@microsoft.com", "user": {"login": "letmaik"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "ec2a8b23862dd3223c071f6396de287bc77281d6", "messageHeadline": "delete test/onnx/simple_thread_pool.h", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 156}}, {"node": {"oid": "e42099480e7c9ad478058b92ee14cf9d662673a9", "messageHeadline": "Clean up code (#1033)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 24, "deletions": 21}}, {"node": {"oid": "46bf6fd1ef4815fc50658618a34aa3afe199e24e", "messageHeadline": "Replace fabs with std::fabs (#1035)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "9029c496b09381f451b66648f7722ad35a43ddf8", "messageHeadline": "Currently download_test_data in build.py supports only the standalone\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "3ac4826268bb6776f57b8a25bf49093914a4dfd0", "messageHeadline": "Don't force the install of a specific numpy version unless explicitly\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "a63598ac5d8b0b8c6ae9a40e6a554f5f2b194cdd", "messageHeadline": "remove old name tables (#1029)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 9, "deletions": 17}}, {"node": {"oid": "0f6b3ea575a7fea39e9f9cec7b964f7777ea7945", "messageHeadline": "Improve the performance of Scatter (#991)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 45, "deletions": 19}}, {"node": {"oid": "cb25fd4d8ad5c7e6eda78a0ef3082a21d41ac687", "messageHeadline": "Updates ARMv7 Dockerfile to use newer CMake (#1017)", "author": {"name": "tmccrmck", "email": "tjmcc01@gmail.com", "user": {"login": "tmccrmck"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "f5dfde33d03605b1a18c4b6ba36e86d5cdffd4f3", "messageHeadline": "generalise strerror_r preprocessor switch (#1026)", "author": {"name": "Maik Riechert", "email": "maik.riechert@microsoft.com", "user": {"login": "letmaik"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "d32d2fbfb7e81e158e06258c22dedd624aaa7ea0", "messageHeadline": "Fix svm runtime (replaces #91, #33) (#101)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 262, "deletions": 162}}, {"node": {"oid": "406770c4846267fa1fb45ac6670451bf50d66ac4", "messageHeadline": "Enable ONNX Runtime Server Model Tests (#1002)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 169, "deletions": 45}}, {"node": {"oid": "f7e57a3d16c31d310e4f47c598cf8b3eac612019", "messageHeadline": "Prune containers and images (#1003)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "9330a42fae6fcc2360e847fa5c406bb1eedc56bc", "messageHeadline": "Optimize upsample (#1008)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 57, "deletions": 34}}, {"node": {"oid": "30f87c7526db14ba18878b435cef22940451365e", "messageHeadline": "Update build instructions based on recent experience. (#1011)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 6, "deletions": 3}}, {"node": {"oid": "340849440751acffe130f62ffe9a20175eae61bd", "messageHeadline": "More C++ API improvements and conversions (#998)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 176, "deletions": 226}}, {"node": {"oid": "c69dff7928151229f510d312877024192665b71e", "messageHeadline": "Implement contrib kernels for Pad (changed interface) and Unique (new\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 766, "deletions": 66}}, {"node": {"oid": "49a2928a5928934f9f8d068c250bb087b897a342", "messageHeadline": "Support string type for split op (#1014)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 148, "deletions": 50}}, {"node": {"oid": "31cbb5d33de775dae852a4b566e3a82cadcc1c0e", "messageHeadline": "Add support for int32_t in CPU provider's Where op (#1009)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 3, "deletions": 1}}, {"node": {"oid": "4e5fdd316798603b675b590dfb4cd2f88f1bc078", "messageHeadline": "Removing one of the generate tranform tests (#1010)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 0, "deletions": 19}}, {"node": {"oid": "c0acb8b6c3b2e3e174627dcb3100009e97c2293d", "messageHeadline": "Allow loading model from in-memory byte-array (#718)", "author": {"name": "Mika Fischer", "email": "mika.fischer@zoopnet.de", "user": {"login": "mika-fischer"}}, "additions": 95, "deletions": 25}}, {"node": {"oid": "af7090c25fb356c755f2c0cecc5275b6a444e95d", "messageHeadline": "Fix server's executor memory leak and add tests (#984)", "author": {"name": "tmccrmck", "email": "tjmcc01@gmail.com", "user": {"login": "tmccrmck"}}, "additions": 108, "deletions": 22}}, {"node": {"oid": "3fb24f6409b7508dcdeaacb8c1293fe4132ebac7", "messageHeadline": "Improve error messaging based on raised issues (#988)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 55, "deletions": 17}}, {"node": {"oid": "b08dbd37cb6861cb761fed3f6e3c47f81ef9b558", "messageHeadline": "Adding Min cuda kernel (#992)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 82, "deletions": 1}}, {"node": {"oid": "2de1f43a40d72ba119c526dcfecfc348d8133765", "messageHeadline": "Move EP docs to the docs folder. (#996)", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "a463f765fc195bfe18a5244de0b4c492a285df75", "messageHeadline": "Minor fix to pass x86 native build (#994)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "2512b0ebeb1b455ac3683cfd791f235f48e0091c", "messageHeadline": "Adding the onnxruntime Dockerfile and instructions (#968)", "author": {"name": "Harry Yang", "email": "huaimingyang@hotmail.com", "user": {"login": "tianchijushi"}}, "additions": 54, "deletions": 0}}, {"node": {"oid": "9a4128efac75de79f891ce1a4cc89c16a6d0f667", "messageHeadline": "Enable model test in Mac pipeline (#990)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 6}}, {"node": {"oid": "c2b412f7be037d309eb92ae58deba1ae43321a5c", "messageHeadline": "Update the ONNX Runtime Server CI pipeline setup (#986)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 78, "deletions": 8}}, {"node": {"oid": "47171a076b81f9a1c924a4e36d99aed18cc5362e", "messageHeadline": "Addressing #956, deprecating old CPU and GPU quickstart dockerfiles (\u2026", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 2, "deletions": 117}}, {"node": {"oid": "926393286183183cf6f113eddb57f31fd93635f2", "messageHeadline": "perf test: fix an index out-of-bound bug (#982)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 17, "deletions": 13}}, {"node": {"oid": "403fb703a0f9edfc53fe17a7f638f8b883ac3904", "messageHeadline": "Avoid using transformers that rely on contrib ops in tests (#983)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "17690355ed242940b550a688c6043fd40bd04440", "messageHeadline": "Support for building ngraph EP on Windows (#978)", "author": {"name": "R. G. Esteves", "email": "rodolfo.g.esteves@intel.com", "user": {"login": "rgesteve"}}, "additions": 352, "deletions": 54}}, {"node": {"oid": "7c514d6fefc009a7e9a6ab728aeba5794e328caf", "messageHeadline": "perf test runner: support NCHW->NHWC rotation (#976)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 204, "deletions": 111}}, {"node": {"oid": "34e8bea487d1d4c3bd90bedb6aa814e723840730", "messageHeadline": "Uncomment mod tests and make infra run them (#971)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 28, "deletions": 14}}, {"node": {"oid": "f2999cf4c2134d4489ccbad2e98a242efa331382", "messageHeadline": "move non_max_suppression source code file to object_detection folder \u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "468de7c8afb7dcdc8be9692675b3d2776a819486", "messageHeadline": "Zhalei/erff (#846)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 1484, "deletions": 1}}, {"node": {"oid": "7e88ca19ee63af33fabbb72a3d2ec5dd6fc7a147", "messageHeadline": "Support the option to disable memory arena in MLDNN provider (#970)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 28, "deletions": 18}}, {"node": {"oid": "8b3c8561ced538c6c2070f5bbe98dd3cd5755522", "messageHeadline": "Avoid to call memory allocation if tensor size is 0 (#972)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "af4a41fd13012c4d2ac60a46b16161f23afd9fbb", "messageHeadline": "change impl of CUDA Greater kernel to avoid data corruption. (#969)", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 11, "deletions": 2}}, {"node": {"oid": "f73ce305e94777544aebb2aa72446124f21b0c9a", "messageHeadline": "C++ wrapper for ABI (#958)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 623, "deletions": 137}}, {"node": {"oid": "7329e99d52078e6cacb0c6e1a7e41868815a246c", "messageHeadline": "Extend IsSupportedOptypeVersionAndDomain to support multiple versions\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 25, "deletions": 27}}, {"node": {"oid": "ab3355f6b4c0e9c5d533ba930f07f6161cc03be0", "messageHeadline": "Add Split op CUDA implementation (#964)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 304, "deletions": 43}}, {"node": {"oid": "f4fd36ee910ff987dd6bca65fd8fc0f4e0dc04e2", "messageHeadline": "merge rel-0.4.0 into master (#959)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 13, "deletions": 9}}, {"node": {"oid": "306453f9d636cd0f47464f75d28111db40feb4fe", "messageHeadline": "fix the link to the script in the doc. fix some error messages (#960)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "0b5f06b0fd16b6ea2e2f77cbcb2a33db5b148f7d", "messageHeadline": "removing LLVM dependency for ubuntu tensor_rt build dockerfiles (#954)", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 0, "deletions": 49}}, {"node": {"oid": "628f4c3aa3bb03fb62fd580a897ec8b6de32e430", "messageHeadline": "Remove unnessary logger from GRU (#951)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 21, "deletions": 12}}, {"node": {"oid": "2c46fff69aa3748095ec9654e7a75d701136422e", "messageHeadline": "Enable gen-doc on windows CI (#716)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 183, "deletions": 601}}, {"node": {"oid": "a02ff0bd06efa65677679ebd34db9c9176b23f83", "messageHeadline": "Fix ngraph build on Redhat Linux", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "6550bb5a529d7c5f34f1609815ab5dcebd880d0b", "messageHeadline": "Add Max Cuda implementation (#950)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 99, "deletions": 3}}, {"node": {"oid": "feab3088fb0c812a82a270841d166e3e44b69db0", "messageHeadline": "Conv(Add|Mul|BN)Fusion as rewrite rules (#863)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 477, "deletions": 478}}, {"node": {"oid": "0ad940027c0b93e971663cdf3fbcc83eb10b7b5b", "messageHeadline": "Use ConstPointerContainer for Node::ImplicitInputDefs() for better co\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 60, "deletions": 48}}, {"node": {"oid": "df513c7fe6eb957d55b85293aeda790ce9ddaf71", "messageHeadline": "Perform shallow clone for nGraph (#938)", "author": {"name": "Tomasz Do\u0142bniak", "email": "tomasz.dolbniak@protonmail.com", "user": {"login": "tomdol"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "7f14326dab4288edb43118f956cfda8b3bbfdc89", "messageHeadline": "nGraph EP documentation (#939)", "author": {"name": "nivas-x86", "email": "43652421+nivas-x86@users.noreply.github.com", "user": {"login": "nivas-x86"}}, "additions": 57, "deletions": 0}}, {"node": {"oid": "1978b3c953ea6218eb1380c63bb31f6494cd9769", "messageHeadline": "Add an HTTP server for hosting of ONNX models (#806)", "author": {"name": "tmccrmck", "email": "tjmcc01@gmail.com", "user": {"login": "tmccrmck"}}, "additions": 5046, "deletions": 3}}, {"node": {"oid": "6f5c28fd3aba0fe7003ae989f1e37715d177cea7", "messageHeadline": "Accomodate missing optional 'axes' when 'steps' is present in Slice o\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 41, "deletions": 24}}, {"node": {"oid": "62b340608c6284a068f2d23bbd181030455c5aa0", "messageHeadline": "Execute one task in current thread (#947)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 12, "deletions": 8}}, {"node": {"oid": "3b0dda0acaacdcf455e232dc7f674b7f96f9bbb9", "messageHeadline": "nGraph: Avoid input and output data copies (#940)", "author": {"name": "nivas-x86", "email": "43652421+nivas-x86@users.noreply.github.com", "user": {"login": "nivas-x86"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "01cd7eaca8ff07e6c3f13827486dee6bd50e0d10", "messageHeadline": "Bump up onnx version (#936)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 4, "deletions": 4}}, {"node": {"oid": "2d00c620051dda714dbb39986216fb0599084119", "messageHeadline": "perf test: support more data types for the TF backend", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 60, "deletions": 6}}, {"node": {"oid": "1b7d1f264562a99dee16d3419aebacb70d02db61", "messageHeadline": "Convert constant folding to a transformer (#866)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 93, "deletions": 66}}, {"node": {"oid": "5bcd77e488e2b5152c397fe91436249b5867172a", "messageHeadline": "Use O_TRUNC when saving ONNX models to prevent possible file corrupti\u2026", "author": {"name": "Sergii Dymchenko", "email": "kit1980@gmail.com", "user": {"login": "kit1980"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "1658efd953d8c7039e8f5555a917a54eaa0260cd", "messageHeadline": "Adding Resize CUDA kernel (#869)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 308, "deletions": 16}}, {"node": {"oid": "ded7eeb03398a0bf978a11a0ac25e01a2b124b89", "messageHeadline": "make builds more robust (#906) (#932)", "author": {"name": "Maik Riechert", "email": "maik.riechert@arcor.de", "user": {"login": "letmaik"}}, "additions": 47, "deletions": 36}}, {"node": {"oid": "8d6114038f7c0d99de6b9b973f617416ea1a2d3e", "messageHeadline": "Adding cuda Greater kernel (#933)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 353, "deletions": 292}}, {"node": {"oid": "f39a8d1f5904a15a809c21b45d27628bc3223ccc", "messageHeadline": "allow users to set graph inputs and outputs fully. (#905)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 90, "deletions": 102}}, {"node": {"oid": "bb588068720c17e24f99402dc458bb45e5c79608", "messageHeadline": "Adding versioned dlls to tar/zip packages (#928)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "93d798b8cae13cb46d15fdc892b5e9de62a490e4", "messageHeadline": "Add locale configuration doc.", "author": {"name": "Dmitri Smirnov", "email": "dmitrism@microsoft.com", "user": {"login": "yuslepukhin"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "71ddba725497127b7d155527340c621ea048b1ed", "messageHeadline": "Disable flaky model test in CUDA build", "author": {"name": "hariharans29", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "861b9fda45c200107a05b7370453f42c9aefdf3b", "messageHeadline": "Add link to build within Nuget package (#926)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "06e0f7e3e7f8f9be4c72e93db495e6d09aa3e532", "messageHeadline": "Minor changes to support inclusion of x86 bits in the Nuget packaging\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 48, "deletions": 15}}, {"node": {"oid": "7a42ffd15fab88ed1337359358ff3703f4cdf82e", "messageHeadline": "Fix in Slice Elimination (issue #885) (#918)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 40, "deletions": 22}}, {"node": {"oid": "90544ed76681af9016d2fa931f486421000ef365", "messageHeadline": "bump version number for release (#911)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "b5ea3973c4d2ea7612a97d95181fdbe32f2766c0", "messageHeadline": "nGraph EP disable new quant tests (#920)", "author": {"name": "nivas-x86", "email": "43652421+nivas-x86@users.noreply.github.com", "user": {"login": "nivas-x86"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "3812c881f70d0e07e045f480cc7fa649088b39e7", "messageHeadline": "Readme updates (#912)", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 23, "deletions": 23}}, {"node": {"oid": "38f1f6943216abbd94d0efbe2b10f3efb1bca99f", "messageHeadline": "Add a temporary bypass of artifacts permission issue (#921)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 17, "deletions": 8}}, {"node": {"oid": "5ed3db914e0c42804da2fedef9d04e5e64a5e3df", "messageHeadline": "Update custom op help (#914)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0cef3b53df93abab6f9ecf31b0352434ef5a6250", "messageHeadline": "Fix scatter assertion offset failure (#913)", "author": {"name": "Bowen Bao", "email": "semisqg@gmail.com", "user": {"login": "BowenBao"}}, "additions": 23, "deletions": 1}}, {"node": {"oid": "8633e9ffda549159626157914c580d37fe6c44a6", "messageHeadline": "Fix a issue that ort will crash if the model has subgraph created by \u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "da0f9bf9a4cf76a1c06c0b864e162eed1990c9de", "messageHeadline": "fixing a bug in resize op (#910)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 5, "deletions": 7}}, {"node": {"oid": "5e3a266709f9ac4873b2a47370b058ca6e917f96", "messageHeadline": "Rashuai/top k0 (#909)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 21, "deletions": 22}}, {"node": {"oid": "0d2181cf85c276b316803c34778d67467783d5d9", "messageHeadline": "Remove parallelfor for certain ops (#908)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 217, "deletions": 117}}, {"node": {"oid": "893b48e92ac2a03edda97d495bc228bc3969b3b3", "messageHeadline": "Implement Mod operator (#900)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 427, "deletions": 35}}, {"node": {"oid": "b8eaa88bd4b04255e53a7ee27b9889ef1997b32b", "messageHeadline": "Migrate ReverseSequence from contrib op to ONNX opset 10 (#896)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 29, "deletions": 169}}, {"node": {"oid": "8ed3eed7b5481cd6f8819035fb5788971cf033ac", "messageHeadline": "Fix ceil_mode not defined for mkldnn pool.cc warning (#907)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "80ac85801656197ad52018baddf84b301500fd4e", "messageHeadline": "Remove OSes/architectures that we don't build on and have no CI for. \u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 19, "deletions": 26}}, {"node": {"oid": "125a77bec4d27680a3d6b61f6218f59fce43d8ed", "messageHeadline": "MaxPool+AvgPool - opset 10 [Dilation and ceil_mode] (#873)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 244, "deletions": 71}}, {"node": {"oid": "3ff97de8dadf2c8387a32e9c8d03502c6df50d2d", "messageHeadline": "Modify roialign to conform with the new onnx spec and take it out fro\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 209, "deletions": 145}}, {"node": {"oid": "73bc09421c1bdafda21402ab168ec397ec9e8eb0", "messageHeadline": "Fix deadlock in parallel executor (#891)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 29, "deletions": 15}}, {"node": {"oid": "ba3b82648ecada7cac9c5bd943cf769386ded014", "messageHeadline": "ng ep update1 (#895)", "author": {"name": "nivas-x86", "email": "43652421+nivas-x86@users.noreply.github.com", "user": {"login": "nivas-x86"}}, "additions": 39, "deletions": 24}}, {"node": {"oid": "95ac7a2f3597b560638667bbdd6030d929bba7a4", "messageHeadline": "Implement separators as regex (#857)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 166, "deletions": 308}}, {"node": {"oid": "1f066d4dc4d35658433bdd7ad0c8f04c79e5a4e6", "messageHeadline": "Update onnx (#893)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 17, "deletions": 27}}, {"node": {"oid": "9d89b23d81bbbf9d90fe5e69d4d90cddfb71dbd0", "messageHeadline": "BatchNorm CPU does not support non-spatial cases - explicitly handle \u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 12, "deletions": 5}}, {"node": {"oid": "d0f846aad5168ff3de50bb2e06c80b99c3da41ea", "messageHeadline": "Tuning GRU performance for batch size >= 2 (#644)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 167, "deletions": 429}}, {"node": {"oid": "80d69515ed8ae11f97c5cb0f005452b3b8e83cef", "messageHeadline": "C API: catch exceptions in OrtCreateStatus (#821)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "11806529d0a12612724f3374721a9c547e8b7f43", "messageHeadline": "Update test data (#864)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 16, "deletions": 29}}, {"node": {"oid": "4b4b585f58cd329a1c7737aa6bcfaf1a47cfd76a", "messageHeadline": "Fix minor bugs in RemoveDuplicateGraphTransformer (#883)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "fb3b63438df7f89932c051fa09b2cf0872139129", "messageHeadline": "Add python api for graph optimization level (#882)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 65, "deletions": 57}}, {"node": {"oid": "b0a37477db118374344e371713c0682c1590df83", "messageHeadline": "Fix memory corruption issue when CPU->CUDA memcpy is involved (#879)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 18, "deletions": 0}}, {"node": {"oid": "7d7627b1acf9686bc224fcec59cc186807c94168", "messageHeadline": "Implement IsInf (#871)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 194, "deletions": 3}}, {"node": {"oid": "0bf12e9dbf7d5ba434dcaa8254a9495b9d3d7bd7", "messageHeadline": "Add option to enable/disable memory pattern back (#872)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 52, "deletions": 13}}, {"node": {"oid": "e8d722003a2e97d1737dbbf555cd3586f70d87b2", "messageHeadline": "Move NMS to Onnx domain (#865)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 83, "deletions": 137}}, {"node": {"oid": "2947e1f9d43d5087080ed8afa623046ce950a4f1", "messageHeadline": "Fix onehot code arm build break.", "author": {"name": "Lei Zhang", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "2879ee8bd2cc7bacf556560c48ead263ed94a047", "messageHeadline": "Fix a few warnings (#762)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "cb69c6575645f1e316c996882e9eb9551908b269", "messageHeadline": "Update MLAS to be able to build standalone again (#874)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 138, "deletions": 254}}, {"node": {"oid": "a4d7052aeb421f18371eab0961f9eaf495c6420c", "messageHeadline": "Add nGraph Execution Provider (#832)", "author": {"name": "nivas-x86", "email": "43652421+nivas-x86@users.noreply.github.com", "user": {"login": "nivas-x86"}}, "additions": 2104, "deletions": 21}}, {"node": {"oid": "7e1edbb9a2cc0dce75940fc7e5695a852a2337e2", "messageHeadline": "Fix a build error in onnxruntime/core/common/threadpool.cc", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 2}}, {"node": {"oid": "d78c340eac0dc251c1e96378fc1dab6e6cf8fdcc", "messageHeadline": "update onnx (#861)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 9, "deletions": 6}}, {"node": {"oid": "b2268a6378cb32b9644d1a7380e7f65c7970ef14", "messageHeadline": "removing specific target framework for c-api test (#860)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "07a4ecbddbeb085b699fb09823990962b025768d", "messageHeadline": "Disable tests for certain models (Cherry pick from 0.3.1) (#842)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 55, "deletions": 3}}, {"node": {"oid": "780aad8fd0b0317ad3bba67d5bd4ba2e205c9b36", "messageHeadline": "Eliminate unused code and data from Linux binaries. (#849)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 10, "deletions": 8}}, {"node": {"oid": "f09a76d669d6d2e93351f304d872a83917b141a1", "messageHeadline": "Don't trigger constant folding in subgraphs (#858)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 26, "deletions": 0}}, {"node": {"oid": "687bac455db2f825283d94c39f8dd55a2ea15289", "messageHeadline": "Convert eigen to a submodule and update it to the latest version", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 43, "deletions": 56}}, {"node": {"oid": "ada90086f727525f48853d9d1e895a0faec4b576", "messageHeadline": "More efficient rule-based transformer (#815)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 198, "deletions": 147}}, {"node": {"oid": "ed0c86cd90b021f9b9491e4755888e8b8413e396", "messageHeadline": "update onnx to fix matmul shape inference (#847)", "author": {"name": "Bowen Bao", "email": "semisqg@gmail.com", "user": {"login": "BowenBao"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "f2694ab526770849e21eeb8ee42c3fed689fea6c", "messageHeadline": "Enable provider unit tests for TensorRT (#802)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 375, "deletions": 276}}, {"node": {"oid": "ff253631b542da6949f05cb16096a0fe21b1e0be", "messageHeadline": "Enable use of session based threadpool. (#854)", "author": {"name": "shschaefer", "email": "stuart@theschaefers.com", "user": {"login": "shschaefer"}}, "additions": 684, "deletions": 440}}, {"node": {"oid": "ac82c1f4835d0861ea32d26f89dcce00f909955b", "messageHeadline": "enable android build (#715)", "author": {"name": "daquexian", "email": "daquexian566@gmail.com", "user": {"login": "daquexian"}}, "additions": 213, "deletions": 42}}, {"node": {"oid": "951c428ee1efc78a92e354e27463946dc1785ced", "messageHeadline": "Simplify the validation in Run call (#850)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 48, "deletions": 91}}, {"node": {"oid": "38a0c0b0a74aa822e54f4beb885fd6ccddcb4543", "messageHeadline": "Fix a bug in perf test runner", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 5}}, {"node": {"oid": "2a2de42bb257c7288103dd96ececebecedfb6c57", "messageHeadline": "Add docker image clean script (#844)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 30, "deletions": 0}}, {"node": {"oid": "f1af493b75830a9b2e68f2795d0dc3bcfb976ed4", "messageHeadline": "Fix some issue in CUDA GRU and ReduceSum (#845)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 209, "deletions": 43}}, {"node": {"oid": "9fb7e98c0b7dcada9eb565cc4ae208d0cf0bf845", "messageHeadline": "fix the allocator type in lru of cuda conv algorithm cache. (#848)", "author": {"name": "Rui Xia", "email": "harrysummerxr@gmail.com", "user": {"login": "harrysummer"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "41dc3130f5b4120ab5cb15734279fb65eee4160b", "messageHeadline": "no need putting initializers (for constant node) into graph inputs. (\u2026", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 100, "deletions": 179}}, {"node": {"oid": "60d71d63b58e5aaf7ec7c12567be5dcf015d254b", "messageHeadline": "Rashuai/onnx test reduce mem (#790)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 72, "deletions": 45}}, {"node": {"oid": "3a8b9a4918b27c50ae4b095e37f1d222f95d6bd2", "messageHeadline": "fix trivial size_t warnings (#843)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 52, "deletions": 52}}, {"node": {"oid": "14d63b5f45e9a22948055e65942faa7988a2df92", "messageHeadline": "generate transformers bug fix (#838)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 55, "deletions": 42}}, {"node": {"oid": "1818835795d846c361688057e0e229768d92cb2f", "messageHeadline": "Adding kernels for Resize op (#809)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 423, "deletions": 290}}, {"node": {"oid": "29ad798c56b25ae7048ce36a4ec1281d876d31ef", "messageHeadline": "Update license - came up during IP scan (#841)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 16, "deletions": 2}}, {"node": {"oid": "07e6dfa7ab93574303109fa3315cd1b4fcad50d6", "messageHeadline": "update onnx and enable tests for qlinearconv (#840)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 9, "deletions": 13}}, {"node": {"oid": "7775551a6fe5aad582550f9a4742734678a90574", "messageHeadline": "Refactor C# and native packaging tests (#825)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 42, "deletions": 133}}, {"node": {"oid": "54e04cb8bb99d749550aa83d577220c00837ad1b", "messageHeadline": "cherry pick PR from 0.3.1 release - enable MSVC static runtime (#837)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "b2658b3594eb54ba7a4d0bf96bc19d90cc708364", "messageHeadline": "Cache CUDNN convolution benchmark results in cuda::Conv kernels (#712)", "author": {"name": "Mika Fischer", "email": "mika.fischer@zoopnet.de", "user": {"login": "mika-fischer"}}, "additions": 149, "deletions": 49}}, {"node": {"oid": "f19d9a490798a5b9769940db238c6119907a8a67", "messageHeadline": "Reduce code size of kernel registration (#833)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 675, "deletions": 645}}, {"node": {"oid": "049ba2d74701f8ce3683b0a0d270112e0374dadb", "messageHeadline": "Exclude tests that fail when contrib ops are disabled. (#835)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "4b4a3599433fa768cf8ffa42f6dd0bc8f66737d4", "messageHeadline": "Exclude unreferenced global data and op doc strings in the opschema o\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 13, "deletions": 7}}, {"node": {"oid": "e999af61b29f98847d3aaaaab77f34dda1341050", "messageHeadline": "bug fix for shape inference (#834)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 1, "deletions": 17}}, {"node": {"oid": "fabdbdc130b81371e5ca61035fd41019c07015b8", "messageHeadline": "Update test retrieval following #828 (#836)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 34, "deletions": 2}}, {"node": {"oid": "6194a92249c895231e1bd5010d7602dd8fdad237", "messageHeadline": "Fix empty input handling in Tokenizer. (#826)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 96, "deletions": 17}}, {"node": {"oid": "2c0b8e965efe00c4767ebdf30834987e83aa21f1", "messageHeadline": "Disable test data local cache in Linux CI pipelines", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 101, "deletions": 8}}, {"node": {"oid": "e493ba2219837ca9a9ac1251b82a0f3b39c38b70", "messageHeadline": "Fix memory leaks in perf test runner", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 44, "deletions": 36}}, {"node": {"oid": "1936d141a7c35bcaa1262de8906fe7ba850b2ab4", "messageHeadline": "Create nightly build for python packages (#817)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 54, "deletions": 15}}, {"node": {"oid": "c55e2de5935391c6864f6fb41c6832bf3c57cfaa", "messageHeadline": "Status class optimizations (#824)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 18, "deletions": 12}}, {"node": {"oid": "b6936e71cbd95187a6dbd507373f0db6551586f2", "messageHeadline": "Avoid postfix iterator increment in a loop in Slice op and some minor\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 71, "deletions": 57}}, {"node": {"oid": "ccf3566c35107f49d34926efa5b5662a753152ea", "messageHeadline": "Register kernel for Dropout (opset 10) for opset compliance  (#813)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 71, "deletions": 4}}, {"node": {"oid": "6577c3dddf51866c7519eb3f5d6620728d685a29", "messageHeadline": "Extract debug symbols in a separate file and strip the binary. (#811)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 45, "deletions": 16}}, {"node": {"oid": "1ff29bfb3df23bfae97edcf857e06afdc1d08a91", "messageHeadline": "Fix x86 calling convention break (#814)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "0741baf867b4c995601d8510552ecc2afcc6d8ab", "messageHeadline": "Update NMS to support max_output_boxes_per_class = 0. NMS will do not\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 26, "deletions": 4}}, {"node": {"oid": "56749a84ee733b995532823d5ecf6e93d2f45fa0", "messageHeadline": "Implement opset v10 changes for Slice operator (#772)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 806, "deletions": 229}}, {"node": {"oid": "53038b33edb45eec059f4a89d44898ca37e1db26", "messageHeadline": "BuildFusedKernelDef uses N^2 algorithm verifying input constraints; s\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 0, "deletions": 4}}, {"node": {"oid": "d17ae5c093323692d088758c4f1eea2b9423b395", "messageHeadline": "MKLML pipeline - update C# and CMake to handle dll dependencies (#810)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "24d80b4bdaeb0f8a0ff831d872813d6fbd6775e6", "messageHeadline": "Add support for BrainSlice execution provider in Python, if onnxrunti\u2026", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 11, "deletions": 1}}, {"node": {"oid": "10b113f144d0e319988bc3c605f9b1200a74a254", "messageHeadline": "update onnx to bring in quantized ops (#808)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 726, "deletions": 1213}}, {"node": {"oid": "4bc3d6027dab99c180b6cb4db956affa75d45aad", "messageHeadline": "Build perf test runner only if onnxruntime_BUILD_SHARED_LIB is ON", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 9}}, {"node": {"oid": "3dcf82a1f94949cef8afb79319f63d62d0ce2044", "messageHeadline": "Disable some flaky tests with CUDA9 (#805)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 18, "deletions": 0}}, {"node": {"oid": "4e3391ef6027fbc258a47ae431cc1d2d4ba9c034", "messageHeadline": "Refactor NuGet to allow arbitrary PackageId names (e.g. Microsoft.ML.\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 28, "deletions": 58}}, {"node": {"oid": "e7090d72024b92ee2f0904dbc8889f54e1d450dc", "messageHeadline": "move all removed exp ops to contrib ops (#786)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 341, "deletions": 219}}, {"node": {"oid": "0d4055def4d9f1605261c1a7c2df7b98740b06d2", "messageHeadline": "Integrate tensorflow into onnxruntime_perf_test tool", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 469, "deletions": 174}}, {"node": {"oid": "9467c5f9674228af4f3075b39ef248f5c87c1834", "messageHeadline": "Update version to 0.3.1  (patch release) (#798)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "ccd7e801a05d9819ba43386b0623c988f3dd5c72", "messageHeadline": "Fix #612, TfidfVectorizer handles empty matrices as an input (#702)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 126, "deletions": 9}}, {"node": {"oid": "39951f35f45d28250770df32d8495104d7427dbc", "messageHeadline": "Use template windows-build-tools-setup-steps.yml in win pipelines (#794)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 19, "deletions": 145}}, {"node": {"oid": "d91555f99e563c33be21435f45d9742f8fba7adf", "messageHeadline": "fix for tensorrt_basic_test not being run. (#792)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 7, "deletions": 2}}, {"node": {"oid": "5cf72030b2069a7a88a1cf24b5f1daef969fdc58", "messageHeadline": "Rename misleading test names in ConvTranspose op tests (#788)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 18, "deletions": 18}}, {"node": {"oid": "571291c323b4086fca00fd1442b422f71a85e27b", "messageHeadline": "build.sh: don't require user to set --use_full_protobuf with --use_te\u2026", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 10, "deletions": 9}}, {"node": {"oid": "cea2a40bf18537e0e6239ed4021ca4fd9aa7edbe", "messageHeadline": "Clean up ExecutionProvider in CSharp (#783)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 5, "deletions": 25}}, {"node": {"oid": "fda1d0dce926f26056da95ff30eb456ffc7cf8cc", "messageHeadline": "Ryanunderhill/ocr custom op (#744)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 326, "deletions": 314}}, {"node": {"oid": "58ef1306d4b8323caad31312d9f77554d585cda0", "messageHeadline": "Copy inputs and outputs directly in InferenceSession::SaveModelMetada\u2026", "author": {"name": "Tao Qin", "email": "48697806+TaoQinMS@users.noreply.github.com", "user": {"login": "TaoQinMS"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "3eddb2d61e5fe3cafb5fa00b8ba1bb9b9557e677", "messageHeadline": "Add optimization level as cmd line arguments (#776)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "36ed91ee9fe599d445a14f6b8ad5783d81100266", "messageHeadline": "CustomRegistry should use composition instead of inheritence", "author": {"name": "utsabsingharoy", "email": "48694199+utsabsingharoy@users.noreply.github.com", "user": {"login": "utsabsingharoy"}}, "additions": 46, "deletions": 9}}, {"node": {"oid": "867e961ee8464fdbec33e306a4294d761b2c3d99", "messageHeadline": "Remove mkldnn_sgemm from math_util.cc", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 36}}, {"node": {"oid": "ffd907116838ec4b47daf051c08f28f1bed258e6", "messageHeadline": "expose graph node name returning non-zero status code (#714)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 18, "deletions": 2}}, {"node": {"oid": "f4021cf30adb4272f48cb13fc1cc1d0c179060ed", "messageHeadline": "fix a minor inaccurate error message", "author": {"name": "Peng Wang (AI FWK)", "email": "pengwa@microsoft.com", "user": {"login": "pengwa"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ef9a4d98cbe774d9ecd1ea8e33821f3eeca61b17", "messageHeadline": "Expose parallel execution option in C# API (#767)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 88, "deletions": 94}}, {"node": {"oid": "43521c0de70396fc3e56bab82ddd047868e5a60d", "messageHeadline": "update", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 13, "deletions": 14}}, {"node": {"oid": "65c50bb25b2f5a77572fae7a3142ecc07088c84a", "messageHeadline": "Create NodeArg for all initializers if IR version is > 3. (#742)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 22, "deletions": 0}}, {"node": {"oid": "2674d9bd8a1d578bbf1da91766e4aa31b4241063", "messageHeadline": "Fix profiling with C API (#710)", "author": {"name": "Mika Fischer", "email": "mika.fischer@zoopnet.de", "user": {"login": "mika-fischer"}}, "additions": 16, "deletions": 1}}, {"node": {"oid": "c35d9797d7ab57293e3216dc49e19201b0b79af1", "messageHeadline": "Add unique identifier in function subgraph (#771)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 9, "deletions": 2}}, {"node": {"oid": "289663fd8561a3ca018418561436a7a95ea1ed26", "messageHeadline": "reformat onnxruntime_c_api.cc", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "290112d614d9400e49d81f56df1fa043db18c627", "messageHeadline": "Update onnx (#761)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 70, "deletions": 57}}, {"node": {"oid": "261a9078a52a6ad68e4772618c64a47da3045841", "messageHeadline": "Remove redundant code in ml::write_scores", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "512cfdd9feba54bd7cad6a7b9033ac707327fd3f", "messageHeadline": "Generalize node removal (#743)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 366, "deletions": 180}}, {"node": {"oid": "7af35ac1e65cc1e7aed30f52e77b42c6ec9a9b3a", "messageHeadline": "Fix two warnings in graph lib", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 4}}, {"node": {"oid": "a1960854717e54381dbd9e8aa509ca0d42c34cb3", "messageHeadline": "Add error messages to FunctionImpl::FunctionImpl", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "adc00ab5fceeb9fdfc7aaf87f98d7ddd0efadb50", "messageHeadline": "ARM32v7 Dockerfile and build instructions update. (#737)", "author": {"name": "Shawn Callegari", "email": "36091529+shawncal@users.noreply.github.com", "user": {"login": "shawncal"}}, "additions": 141, "deletions": 23}}, {"node": {"oid": "8bc532bfb9e9eee9608d5c52e68e253a68b25fbb", "messageHeadline": "update onnx and add removed experimental ops to contrib ops (#723)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 91, "deletions": 5}}, {"node": {"oid": "6e9ed17adc28a4e62516d24b8c85f2109cb4cfda", "messageHeadline": "Fix a shape inference bug for FusedConv and MaxpoolWithMask (#748)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "fc26b24138de9cf567b2d8c8a5397637773c36e8", "messageHeadline": "onnx_test_runner: use c api for loading non-tensor test data (#751)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 446, "deletions": 421}}, {"node": {"oid": "0cd4a9381c4705071ea1a21026ecbe7faf820094", "messageHeadline": "Handle seqence length == 0 for RNN (#749)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 203, "deletions": 23}}, {"node": {"oid": "2dbce4ebcfe9b7657d2e6aacbf20ef676a4c3495", "messageHeadline": "csharp api for graph transformers (#741)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 40, "deletions": 7}}, {"node": {"oid": "06888437dd4270b07531253673460bb26e9630ee", "messageHeadline": "Update onnx-tensorrt submodule to master (#753)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "acc8ac58d217c1f2d571d35e7b3e947f84431456", "messageHeadline": "Fix C-API sample. Update Issue template. (#750)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "afe3aae29fb0d24322824276bb9e76c8d838e8b6", "messageHeadline": "Support empty tensor concats in Concat op (#735)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 84, "deletions": 22}}, {"node": {"oid": "7d47cd39b62f10aa64b368b322aed19b58be3303", "messageHeadline": "Change bind2nd to bind (#747)", "author": {"name": "Rui Xia", "email": "harrysummerxr@gmail.com", "user": {"login": "harrysummer"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "b1115f49cd0cb516964f06142896188d7c6d0688", "messageHeadline": "Update NMS to compatible with both TF & Pytoch models (#636)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 392, "deletions": 293}}, {"node": {"oid": "f4b47ad9f693ea86909dd820bed8ba68aeabaafa", "messageHeadline": "Move call to log and fmaxf outside of inner loop. (#745)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "40839f1f845b635013e6f10cb0c6ec1d6bc5fbeb", "messageHeadline": "Enable multiple session runs for TensorRT (#724)", "author": {"name": "stevenlix", "email": "38092805+stevenlix@users.noreply.github.com", "user": {"login": "stevenlix"}}, "additions": 35, "deletions": 25}}, {"node": {"oid": "6df54f02852772c99c406b906657953bf3482afc", "messageHeadline": "Eable some onnx test cases (#700)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 3, "deletions": 16}}, {"node": {"oid": "7f96c7f0281a853fc9627dc050445b76fd3e0377", "messageHeadline": "Use forward_inference to speed up convolutions with MKLDNN (#731)", "author": {"name": "Mika Fischer", "email": "mika.fischer@zoopnet.de", "user": {"login": "mika-fischer"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "e6a2bdfacdb584bc77d51076bb49d852febfdeb6", "messageHeadline": "Handle incorrect perm data in Transpose op gracefully  (#739)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 18, "deletions": 3}}, {"node": {"oid": "667fa39551cc3978bf1b06614620a8c437a8df14", "messageHeadline": "fixing spacing for TensorRT documentation (#645)", "author": {"name": "Prasanth Pulavarthi", "email": "prasantp@microsoft.com", "user": {"login": "prasanthpul"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "73fc91dc59d64a50af6075efbea72dbc0eeacc38", "messageHeadline": "Fix preFast native rules warnings (#682)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "b9b6e3abcb158a05ee42f31f0da7daccd8e6ef4f", "messageHeadline": "ReverseSequence contrib op (#728)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 528, "deletions": 33}}, {"node": {"oid": "333171f602f45a5b7ce7222d96905342d9f5abe6", "messageHeadline": "Bug fix for Range op (#734)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 43, "deletions": 4}}, {"node": {"oid": "e643ce0e08e6245a609f04b23d931eff581f978b", "messageHeadline": "Fix inconsistent dimension data type in C-API (#726)", "author": {"name": "Ahmad El Husseini", "email": "49005003+ahmadhusseini@users.noreply.github.com", "user": {"login": "ahmadhu-MS"}}, "additions": 25, "deletions": 25}}, {"node": {"oid": "165657ee1a81d1b2820b63628ae2e60d81c88f21", "messageHeadline": "Refactor the /Qspectre cmake code (#736)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 13, "deletions": 6}}, {"node": {"oid": "6cbf5bcb045ac748ed0bdca01a446a8974dc8aa4", "messageHeadline": "[Minor] Enable pybind in mac build (#732)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 2, "deletions": 1}}, {"node": {"oid": "36f5d008de85bdf4ce263e97f2689a802e33fe6d", "messageHeadline": "Add /Qspectre compiler flags (#671)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 7, "deletions": 0}}, {"node": {"oid": "f6a77617c112cac2d9716334e833b8918adbe07d", "messageHeadline": "update test data", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "deaea702ffa01b06227153421ccf658a0e480487", "messageHeadline": "Bump up cmake_minimum_required to 3.13 (#722)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 7, "deletions": 7}}, {"node": {"oid": "f299104a19f5506657b0a367cbeca6326dc395e1", "messageHeadline": "Enable constant folding in L1 transformers (#720)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "fb2a44f6423ac5d4f8af9105af82f00c99ae3a46", "messageHeadline": "Remove header_files_test.cc", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 88}}, {"node": {"oid": "c35b605b8d2c6224e6693bfcbad103d303c46c71", "messageHeadline": "Support updated opschema with functionbody (#640)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 74, "deletions": 45}}, {"node": {"oid": "83ae641425eb2ef1aec03924cb9e4064a4cd2f68", "messageHeadline": "add documentation for custom ops (#708)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1699, "deletions": 0}}, {"node": {"oid": "77b981824aed851442a3bed33525d27adb525111", "messageHeadline": "fix graph transformers and refactor tests (#696)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 465, "deletions": 339}}, {"node": {"oid": "a872ba7894d52d89f844df6570ba8f305d8b6b72", "messageHeadline": "Convert Unsqueeze elimination to rewrite rule + improvements in graph\u2026", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 355, "deletions": 348}}, {"node": {"oid": "a28b42a42cab28c1292b6e7862aaa8ef499f1784", "messageHeadline": "Fix path_lib.h for Mac and refine #include in InferenceSesssion.h (#711)", "author": {"name": "Tao Qin", "email": "48697806+TaoQinMS@users.noreply.github.com", "user": {"login": "TaoQinMS"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "bcf1ce94be8416ca2577a3b213729f3c2913ba79", "messageHeadline": "Provide an option to disable contrib ops. (#707)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 57, "deletions": 12}}, {"node": {"oid": "39fb68b7612c0f1bf8a2dc888be77f60eee9786a", "messageHeadline": "Refactor InferenceSession class (#654)", "author": {"name": "Tao Qin", "email": "48697806+TaoQinMS@users.noreply.github.com", "user": {"login": "TaoQinMS"}}, "additions": 819, "deletions": 859}}, {"node": {"oid": "c8f1da28c4975d19a89ac2fc45715bb4cd975f2d", "messageHeadline": "tile op: make implementation type-agnostic (and support a few more ty\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 167, "deletions": 89}}, {"node": {"oid": "6497f0c133f1e3520b34dc31603f499e90d2a24f", "messageHeadline": "build python only when onnxruntime_BUILD_UNIT_TESTS is ON (#694)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 7}}, {"node": {"oid": "21dacdd4d67458fe54ac05c15b58bb7a96a69f25", "messageHeadline": "fix epilogue code (#695)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "179afe1594f08e258085de4a030537dd0a967b9a", "messageHeadline": "Remove onnxruntime_USE_PREBUILT_PB (#692)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 45}}, {"node": {"oid": "a26696fb0e3b3d7f911dc60d037cf06df98674ed", "messageHeadline": "Enable LTO on Linux", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 56, "deletions": 24}}, {"node": {"oid": "12ecd77ffbb6fa1990a030f3c034e728938480cb", "messageHeadline": "Update README.md (#686)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "af389593be91c70b498fba1e4976f8170fcb4328", "messageHeadline": "Add Windows CI pipeline for TensorRT (#687)", "author": {"name": "stevenlix", "email": "Li.Steven@microsoft.com", "user": {"login": "stevenlix"}}, "additions": 20, "deletions": 18}}, {"node": {"oid": "21dde6fd16a6fe96a6ac4d84322ff5e7e93ce6f3", "messageHeadline": "Clang build failure in test (#683)", "author": {"name": "utsabsingharoy", "email": "48694199+utsabsingharoy@users.noreply.github.com", "user": {"login": "utsabsingharoy"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "8d782582f4a5f863fecc8ad0368cb36bc795207c", "messageHeadline": "fix build_wheel option (#684)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 4, "deletions": 1}}, {"node": {"oid": "25f45cb2db8d8cb27041dfa490ef78ac37addfea", "messageHeadline": "Introduce Rowwise/Colwise Sum to math util (#656)", "author": {"name": "Sherlock", "email": "SherlockNoMad@users.noreply.github.com", "user": {"login": "SherlockNoMad"}}, "additions": 32, "deletions": 0}}, {"node": {"oid": "600dc9ecc5c6140b43a3cc70c747996e6e173a2e", "messageHeadline": "Remove licenseurl and add licensefile, to fix issue 664 (#669)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 12, "deletions": 6}}, {"node": {"oid": "9b0d56dbedd70f8282b61e2a93e85f81da261551", "messageHeadline": "Fix a warning in GraphTransformerManager", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "cd52431b8fe3fe1a3f32039f46ed9f809b968e7c", "messageHeadline": "Custom op interface to the C API to remove shared library dependency \u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 58, "deletions": 21}}, {"node": {"oid": "6c40aed95c01006372749358abbf03c0c02358e7", "messageHeadline": "Rashuai/build x86 (#676)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "5d452b3029e9638a290ce7aa9f97d659e0accecc", "messageHeadline": "Use protobuf-lite to reduce onnxruntime.dll size. (#639)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 1124, "deletions": 247}}, {"node": {"oid": "a624e1091ef2d2b7bda329165cedf12d96475fd4", "messageHeadline": "fixing clang build failure", "author": {"name": "utsab", "email": "utsingha@microsoft.com", "user": {"login": "utsabsingharoy"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0995e853fabd885de17cbe579c645ed219fb6331", "messageHeadline": "Rashuai/unify version (#653)", "author": {"name": "RandySheriffH", "email": "48490400+RandySheriffH@users.noreply.github.com", "user": {"login": "RandySheriffH"}}, "additions": 86, "deletions": 1}}, {"node": {"oid": "a3499083dae79c5737c2d31d9512068f89cb48e1", "messageHeadline": "Add iterator traits aliases to ConstPointerContainer::ConstIterator (\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 27, "deletions": 3}}, {"node": {"oid": "2f1c3028b7c63c424e055b5f9d7137114ab96314", "messageHeadline": "add capi to set graph optimization level (#657)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 78, "deletions": 30}}, {"node": {"oid": "819457dd45976a39e26b6f564d104fa9790cd9c5", "messageHeadline": "Added netframework test (#658)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 9, "deletions": 3}}, {"node": {"oid": "17af8e9ba78526836ab99d1635ae270a0eba8d8c", "messageHeadline": "Add subgraph check/update to node removal logic. Fix a few minor issu\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 349, "deletions": 26}}, {"node": {"oid": "c3666472628e58a606a63944934efb3ff9edc05f", "messageHeadline": "Add missing probit function for treeregressor (#619)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 21, "deletions": 28}}, {"node": {"oid": "9e323901b20b6d4ea3d4e79dfb72eea985c9e842", "messageHeadline": "Upgrade mkldnn to 0.18 (#650)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 14, "deletions": 9}}, {"node": {"oid": "1aa24cbbf3da60053f98c37aee5505314ea2a9d4", "messageHeadline": "executable size reduction: cleaned up slice op to get savings (#621)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 142, "deletions": 163}}, {"node": {"oid": "3f52de07c7b844ecfdd6310d705d83513402b12d", "messageHeadline": "Add missing include to status.h", "author": {"name": "Casey Carter", "email": "cacarter@microsoft.com", "user": {"login": "CaseyCarter"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "4cc7121368a033ec0d2925c0dfcd6de1774fdd9c", "messageHeadline": "Fixes #626, remove posix option for regular expression in Tokenizer o\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 21, "deletions": 1}}, {"node": {"oid": "da9af592d973620f15914b9b24c4d02592d69ad2", "messageHeadline": "Remove OrtAppendCustomOpLibPath (#642)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 13, "deletions": 239}}, {"node": {"oid": "481eb971ec1adfbef8d10b39977bdffd5467c273", "messageHeadline": "graph transformers update (#608)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 579, "deletions": 178}}, {"node": {"oid": "541b3149ddc190d8d6617211181d469bece1d1d1", "messageHeadline": "Minor fix to disallow an execution provider registering a nullptr.  T\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 8, "deletions": 1}}, {"node": {"oid": "971058fc38b24c8582da032f1421840c8506046f", "messageHeadline": "Avoid copy of pre-existing value to subgraph output (#637)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 69, "deletions": 34}}, {"node": {"oid": "14d9a2bdc78a66d68361c6e275b782d004720728", "messageHeadline": "Remove an unnecessary check in GetDirNameFromFilePath function (#616)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 9}}, {"node": {"oid": "8faccfcd9278deb1b4d2e156369c978c1963c513", "messageHeadline": "Remove parallel logic for backward and forward GPU (#641)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 34, "deletions": 72}}, {"node": {"oid": "4c2b1c30186aa00edbd5fcdecef96cba5846b22f", "messageHeadline": "Rework Transpose as a generic type agnostic implementation (#561)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 228, "deletions": 41}}, {"node": {"oid": "4bd84632280932465acb640af63614f3825db93a", "messageHeadline": "Update docs (#633)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 16}}, {"node": {"oid": "bdc2bbb2070ee92f7588ebfa4c19fce35e481f9a", "messageHeadline": "Build details for TensorRT execution provider. (#632)", "author": {"name": "manashgoswami", "email": "manash_g@yahoo.com", "user": {"login": "manashgoswami"}}, "additions": 78, "deletions": 11}}, {"node": {"oid": "a362c3bbdf103dbae8dc872ee811640648aeb881", "messageHeadline": "Fix node def replacement by checking memory location (#623)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 27, "deletions": 4}}, {"node": {"oid": "5c09aa863c866dbdb68b979a37de1936b8cc1f72", "messageHeadline": "python binding test updates for TensorRT (#635)", "author": {"name": "stevenlix", "email": "Li.Steven@microsoft.com", "user": {"login": "stevenlix"}}, "additions": 16, "deletions": 8}}, {"node": {"oid": "639aa6d97ab03beb77789ba51f7065dfdd9ffb70", "messageHeadline": "Update run_dockerbuild.sh (#631)", "author": {"name": "stevenlix", "email": "Li.Steven@microsoft.com", "user": {"login": "stevenlix"}}, "additions": 1, "deletions": 12}}, {"node": {"oid": "71b64459678036005e296a5a6d2db862e0c32489", "messageHeadline": "Large Model: Support offset (#615)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 110, "deletions": 58}}, {"node": {"oid": "e8b0ae892380972ed695e1bd455f3534a9523f03", "messageHeadline": "Trt execution provider (#382)", "author": {"name": "stevenlix", "email": "Li.Steven@microsoft.com", "user": {"login": "stevenlix"}}, "additions": 1339, "deletions": 61}}, {"node": {"oid": "37f7ed156e96b9d9479054723333b686aa75f6cc", "messageHeadline": "renaming utils namespace related to graph operations to graph_utils (\u2026", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 34, "deletions": 34}}, {"node": {"oid": "fba98bb4de71e205cd202c2bb36a179e835de2db", "messageHeadline": "Rework tokenexp to match tokens instead of separators. (#617)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 58, "deletions": 95}}, {"node": {"oid": "2ae83c580c9cf88f0f416596cce5309d8aa5d5bb", "messageHeadline": "Constant folding (#168)", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 287, "deletions": 62}}, {"node": {"oid": "ab734ec5a64b364fcf8aff3d91eea860887afc42", "messageHeadline": "Fix a bug in ExternalDataInfo", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 4}}, {"node": {"oid": "cfb08c48485977d56ca95af4869ed5c4b071e59a", "messageHeadline": "TopK op: Promote onnx to a newer commit and handle changed TopK spec \u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 268, "deletions": 101}}, {"node": {"oid": "7dd9bc4d78efef5964c8559217afbcc7c3aa417a", "messageHeadline": "Restore changes removed by PR #571 (numpy version) (#603)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 17, "deletions": 12}}, {"node": {"oid": "eab1f5463cf4b8a6019f38d630226e4e6c1be2cc", "messageHeadline": "Disable benign warning that shows up in Windows cross-compiled ARM bu\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 17, "deletions": 14}}, {"node": {"oid": "5bb842538dbdd231daf10fa5350f0f9ba5dc59db", "messageHeadline": "sync onnx and maintain old version history for removed exp ops (#588)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 122, "deletions": 5}}, {"node": {"oid": "c6d39b60cd758a3311f0a59bba68dbfd1130df72", "messageHeadline": "remove OnnxTransformer (#554)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 0, "deletions": 159}}, {"node": {"oid": "bf43ac41aa09db66f96c7e339c17d1be245ba465", "messageHeadline": "fix version number for tarball packages (#600)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 60, "deletions": 10}}, {"node": {"oid": "b452151bafe49da332aaeab639ef0318fc1ca28d", "messageHeadline": "Rashuai/restore yml2 (#604)", "author": {"name": "Randy", "email": "randysheriff@gmail.com", "user": {"login": "RandySheriff"}}, "additions": 162, "deletions": 0}}, {"node": {"oid": "3f507fa9f812e92dfcd0bf0b050dd8311fe1ada8", "messageHeadline": "fix a shape inference bug. (#605)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "de9f1ff1ff8be487b7521bea80b688e4d4e69bf2", "messageHeadline": "Add new C function OrtOnnxTypeFromTypeInfo (#585)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 44, "deletions": 3}}, {"node": {"oid": "f048fc5fb0093e128cc7b2cd8614edb8743e6dae", "messageHeadline": "cross compile x86 linux (#562)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 1067, "deletions": 177}}, {"node": {"oid": "3ef273b84b54cb3327bbc55c2977ab9210d4cf01", "messageHeadline": "Support memory mapping on Linux", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 364, "deletions": 354}}, {"node": {"oid": "7218be4f1f38d98469551bce7d865735c27e10b1", "messageHeadline": "support protoc version < 3.2.0 (#594)", "author": {"name": "liqunfu", "email": "liqun_fu@hotmail.com", "user": {"login": "liqunfu"}}, "additions": 17, "deletions": 0}}, {"node": {"oid": "59cfafbb9e4d5c31856a78a1e71514725bca3df9", "messageHeadline": "Restore tolerance fix and make test_gru_seq_length_cpu run (#595)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 5, "deletions": 7}}, {"node": {"oid": "02ad7daa8b0ad4251231f2ade9bc054f1f54234f", "messageHeadline": "Add component detection (#592)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 20, "deletions": 0}}, {"node": {"oid": "93bbb42e7577776be4323509b13a986cb6516ca4", "messageHeadline": "Change the way of linking dl lib", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 6, "deletions": 8}}, {"node": {"oid": "b183df71430fedcb642cfb75ad7e8bee6be5ed9d", "messageHeadline": "Remove InsertFusedRules from CPUExecutionProvider (#573)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 22}}, {"node": {"oid": "867eda52621b7e1009732d7f78e5603ecccabb7e", "messageHeadline": "Support Windows cross-compiling for ARM(64) in ORT build scripts (#549)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 181, "deletions": 10}}, {"node": {"oid": "6136efc0c039f65c37da14dba06106c64b874cf5", "messageHeadline": "Promote to ONNX commit that has StringNormalizer (#499)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 56, "deletions": 107}}, {"node": {"oid": "530748ad5c29cc433e82ff0e41d13463dd71ab08", "messageHeadline": "Fix the broken link.", "author": {"name": "Weixing Zhang", "email": "wezhan@microsoft.com", "user": {"login": "weixingzhang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d23edeb8c5835225ac427110947d610a99393f44", "messageHeadline": "Fix a warning in env.cc (#568)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "b7298a6deb0e29a7c2886314d4983a515e840985", "messageHeadline": "Add python script to export subgraphs from Scan/Loop/If nodes in an O\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 52, "deletions": 0}}, {"node": {"oid": "fb6f0910739f9aefd3c3b7aaf9a8d182f210e11e", "messageHeadline": "Create derived class from onnx BackendTest so we can use the same tol\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "b742c3a965691065b1ec43ecb1780e9aa825aa3b", "messageHeadline": "Fix for Scan 8 bug where the sequence length was > 2 (#580)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 20, "deletions": 12}}, {"node": {"oid": "c785fe26d377e84a651f9616b4824ab31dd1e37e", "messageHeadline": "Update CreateExecutionProviderFactory_BrainSlice() definition.", "author": {"name": "Jesse Benson", "email": "jesseb@microsoft.com", "user": {"login": "jessebenson"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "44cb4c4f74ba2d92f74b0ee53fab115952d1ed51", "messageHeadline": "Handle x86_64 -fPIC relocations correctly (#566)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 27, "deletions": 17}}, {"node": {"oid": "9fa7b570da1b7e3f042c5ae65e39bbc801e6e966", "messageHeadline": "Fix publishing of Linux and MacOSX artifacts. (#579)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "1180c600354a800df0140a6a148c030140762d03", "messageHeadline": "Add link to custom ops (#576)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "51273b48f6820d0a0b04265924ddf1f2c7dac2f3", "messageHeadline": "Use cuda9 1 in c api packaging (#571)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 46, "deletions": 27}}, {"node": {"oid": "3cd448e05a058302711b4530ef759a9efb2d4841", "messageHeadline": "Fix: IExecutionProvider::GetCapability returns redundant subgraphs", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "e272feb60dd5404b41fdf3ac31e422dd7b318b7c", "messageHeadline": "Readme updates (#570)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 15, "deletions": 13}}, {"node": {"oid": "a79c09388f79626f44de881777735b7f9aab7df3", "messageHeadline": "Fix GPu package testing for CAPI (#569)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 187, "deletions": 1}}, {"node": {"oid": "d40a9f894f7161ce8671a8678b282c7af15b12f8", "messageHeadline": "Enable Component Detection (#559)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 23, "deletions": 136}}, {"node": {"oid": "b4ffcf8258cc7bcd4a97303efd53d6e6c01904d0", "messageHeadline": "Fixes #31, add option numpy_version, skip_keras_test to the parser of\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 18, "deletions": 16}}, {"node": {"oid": "4635bcc62461aba1208bfe1acd44f5040107f5f2", "messageHeadline": "Updating C_API end-to-end test and user samples (#564)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 270, "deletions": 157}}, {"node": {"oid": "b68079fe5d334ab525cc80a901efdc5e4dee4833", "messageHeadline": "Support int32_t for Split op (#563)", "author": {"name": "David Fan", "email": "30608893+jiafatom@users.noreply.github.com", "user": {"login": "jiafatom"}}, "additions": 72, "deletions": 44}}, {"node": {"oid": "af9c554dd354e897cf2fd64714889c4a0639496b", "messageHeadline": "Ryanunderhill/custom op (#550)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 441, "deletions": 50}}, {"node": {"oid": "0b143d07036baae17bc9014ad971c96b2171247b", "messageHeadline": "Fix parentheses and commas (#560)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "b247fced3bb0fa2b5623d864d31a12e55983d7b1", "messageHeadline": "Linux and MacOS C api packaging (#555)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 206, "deletions": 95}}, {"node": {"oid": "98a9c0e715a59f2aabe150498202738ff7444184", "messageHeadline": "Fix json error (#557)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "0e65bfe7ae8ceaf80ccc1c96d09a76825c17aa03", "messageHeadline": "Remove caching from InferenceSession::Run (#547)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 16, "deletions": 220}}, {"node": {"oid": "54eeb7394afe9040277e2083c5647f9418997279", "messageHeadline": "Implement Scatter. (#523)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 346, "deletions": 4}}, {"node": {"oid": "cf41f76d79595fb766444b26d96db3759965e4f9", "messageHeadline": "Fix some warnings (#551)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 24, "deletions": 25}}, {"node": {"oid": "714d4100bd04434e1b3fd75d365cc6374ef21e7c", "messageHeadline": "Update documentation to include openmp dependency. (#545)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 20, "deletions": 15}}, {"node": {"oid": "8e0fff7b8d84cb3485e151aef144819475f12ada", "messageHeadline": "Support large model(>2GB) (#520)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2452, "deletions": 2201}}, {"node": {"oid": "a4a459477a78bc57c476c1979ca04eed7c377fd9", "messageHeadline": "Windows packaging build pipeline for C-api packages (CPU and GPU) (#535)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 280, "deletions": 16}}, {"node": {"oid": "3a2f6c69646df66fe82c919cd956a2034a418a0e", "messageHeadline": "Update cgmanifest and TPN  (#529)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 44, "deletions": 116}}, {"node": {"oid": "94bd74190ab383fda0dddcdbfc022016ef1d28ce", "messageHeadline": "Revert to cuda 9.1 for package release (#546)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "47a9abd2125770854ac7964e5e881b0357fae296", "messageHeadline": "minor change for validating input types (#532)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 53, "deletions": 77}}, {"node": {"oid": "bfa21c0b9f22d1c5564d3c47f202306e1b1020d4", "messageHeadline": "Re-enable the tests disabled during C API renaming", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1288a8caed7f473d42a3b2cc6f9d2c31011d328d", "messageHeadline": "Initial check-in to support non-tensor (sequence/map) types (#527)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 373, "deletions": 55}}, {"node": {"oid": "1d3fcc525a77cf96e4ba90d876dde966c8748a32", "messageHeadline": "deps: update onnx to a newer commit and update test exclusions (#542)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 21, "deletions": 5}}, {"node": {"oid": "30be4c1ef843f15af171066a3dc2c52b8a187a83", "messageHeadline": "Fix a bug in test data (#539)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 25, "deletions": 25}}, {"node": {"oid": "f5dfbba6556d04c515ac7c22743b52354f5dd1bb", "messageHeadline": "Clarify numpy version requirement (#537)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 6, "deletions": 4}}, {"node": {"oid": "7aba5fe5235868af456b68b86b488651c774dcf5", "messageHeadline": "perf_test_runner: support specifying thread number (#538)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 8, "deletions": 3}}, {"node": {"oid": "4c684a133a5478085631761aebfaf6ce1b02c7ca", "messageHeadline": "bump up version to 0.3.0 (#536)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "8a59287c46499fa026f2ae0d275d80732ce4c610", "messageHeadline": " Create OptimizerExecutionFrame for graph optimization (#526)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 326, "deletions": 6}}, {"node": {"oid": "290c4728391bbba1d8ccddb293b630e07d216432", "messageHeadline": "remove mkldnn from linux packages (#533)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a697e0b7101a01577798e66d62383b79efce6f5f", "messageHeadline": "Implement Shrink operator (#485)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 302, "deletions": 4}}, {"node": {"oid": "9fb80ea9275515e135e89df7379b06e818cd538f", "messageHeadline": "Fix cast optimizer (#524)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 8, "deletions": 0}}, {"node": {"oid": "2e6ec07d9abca13e9f683418f2ea6e9f9b610cd9", "messageHeadline": "Check some return status values that were ignored and add logging of \u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 32, "deletions": 11}}, {"node": {"oid": "6c7099a18e148a9edbcde22a04a3e36ee5865485", "messageHeadline": "Break dependency on SessionState for ExecutionFrame and OpKernelConte\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 524, "deletions": 434}}, {"node": {"oid": "dfa21af302c75b5adacf3259261641e69c64b1f0", "messageHeadline": "Update C API to allow user to enable caching of feeds and fetches inf\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 32, "deletions": 28}}, {"node": {"oid": "96d521186984a5a763cd7e4a83cc5c2abf17b091", "messageHeadline": "Update cgmanifest (#502)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 54, "deletions": 1}}, {"node": {"oid": "6e79c93dd289d0f5389d28fdc06911815d4361e6", "messageHeadline": "replaced onnx namespace with ONNX_NAMESPACE (#519)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 67, "deletions": 69}}, {"node": {"oid": "bdf72ad5e3d30607a94b65c6690fa4749cc44551", "messageHeadline": "Update C-API with working example (#503)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 161, "deletions": 2}}, {"node": {"oid": "0b841c3bd410b6c0dbf888b26e59e33df7898a67", "messageHeadline": "Fix various issues with murmurhash. (#514)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 53, "deletions": 19}}, {"node": {"oid": "f9bae489bdd4e662495cf6e9d27872e775b2218c", "messageHeadline": "cleanup extra header from c api and sanitize C api test (#517)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 80, "deletions": 70}}, {"node": {"oid": "668fcf22d88524ce9b0330b3432147767bd4b3ee", "messageHeadline": "Update InferenceTestCapi.cpp (#516)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "2f1e883c71ae35d569a663ff3711b969674769cc", "messageHeadline": "Don't use mkldnn and tvm for release pkgs. (#511)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 7, "deletions": 6}}, {"node": {"oid": "4c7fd499496c20e1a2ad9c1588c607e5ff2ccf2e", "messageHeadline": "Add a few benchmarks (#512)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 59, "deletions": 8}}, {"node": {"oid": "5171e8b1297918ce7f3837c6aae12fbb35fc0730", "messageHeadline": "Make IExecutionProvider::Type return const std::string& instead of a \u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 24, "deletions": 35}}, {"node": {"oid": "d6a70470dda0387e4f600d20c1fb24640635dc96", "messageHeadline": "Update build.md to say CUDA 10 should work with any VS2017 version. (\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 2}}, {"node": {"oid": "9d14cbdb1a1261e7f8bcd46c3d8e9132e529f9fe", "messageHeadline": "Throw friendly error message when Linux distribution has libc version\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 28, "deletions": 0}}, {"node": {"oid": "62bbe8de40112b7e7b622597f59aed1f1d7c1605", "messageHeadline": "fix downcast errors in 32-bit build (#501)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c2b8ac01544b5b2012ff60009aefb59e53606070", "messageHeadline": "MatMul op: Support new integer types and double type as part of opset\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 408, "deletions": 114}}, {"node": {"oid": "b69c834c06ed93e867d9df17cb1cc77056445277", "messageHeadline": "Optimize graph partition", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 589, "deletions": 262}}, {"node": {"oid": "0e687a2c901ccd4a0920cdea41aed2247711be97", "messageHeadline": "Implement tokenex regular expression matching and add tests. (#480)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 442, "deletions": 89}}, {"node": {"oid": "b02c1d80d4bb7ccb2036a4af97f98eb919cc758c", "messageHeadline": "Fix an SAL annotation in the C API", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "fc7185f06021cfd05cc7a86a0bff9dac0d9e3309", "messageHeadline": "Various optimizations to reduce the setup and device copying cost out\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 1505, "deletions": 492}}, {"node": {"oid": "4408342f0b36c15b5ac74bcf09744539b7ac1abb", "messageHeadline": "Fix a issue for cuda reduce Max Min (#474)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 25, "deletions": 14}}, {"node": {"oid": "9bc6503463762b3f2c67306274060ef361327f1b", "messageHeadline": "Support non-tensor types in the C API. (#489)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 735, "deletions": 4}}, {"node": {"oid": "da1cf8fff0600369c19c8cf371ff247cea6a93a9", "messageHeadline": "Remove exclusions for Sign operator model tests. (#490)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "54acfc0432bf8fb69a6fe35476915db54a389490", "messageHeadline": "TPN update and link fix (#483)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "ee702bd2889a8c8b42a5c606b79a7f0137f3f32c", "messageHeadline": "patched the logic of removing the ._*.onnx file, in case it comes in \u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 12, "deletions": 4}}, {"node": {"oid": "1f1dcc352fc6000c03993d9621b9e05b4e6f4766", "messageHeadline": "Add Native C API test from NuGet (#481)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 270, "deletions": 5}}, {"node": {"oid": "2a9a924c23cac9183e60d251fc91ec2d9835b5c6", "messageHeadline": "Add float16 support for fusion (#476)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 146, "deletions": 0}}, {"node": {"oid": "9add0e9a9f295c4f7570a2090b15649fc379f1d5", "messageHeadline": "random generator to continue generate random numbers (#477)", "author": {"name": "liqunfu", "email": "liqun_fu@hotmail.com", "user": {"login": "liqunfu"}}, "additions": 35, "deletions": 12}}, {"node": {"oid": "f6ffa1280a2d8b75114f74b1a831b5d055708a96", "messageHeadline": "Updated endtoendtests to not copy model files (#479)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 5, "deletions": 4}}, {"node": {"oid": "62532ec1b0de9e0d9abda56ab21f2a6c49f0b087", "messageHeadline": "Minor cleanup in TopK operator (#478)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 1, "deletions": 11}}, {"node": {"oid": "f2510127a2431fbd53bd8150c2e81a1cfe92ac21", "messageHeadline": "Optimize pad performance (#472)", "author": {"name": "ybrnathan", "email": "7902510+ybrnathan@users.noreply.github.com", "user": {"login": "ybrnathan"}}, "additions": 107, "deletions": 25}}, {"node": {"oid": "0a23d232668f31d5a510856844946021fab88ba1", "messageHeadline": "Initial implementation of Where op. (#412)", "author": {"name": "edgchen1", "email": "edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 370, "deletions": 23}}, {"node": {"oid": "d05b74b1b7ee6c066c11f05d79dbab8f96c2a4fb", "messageHeadline": "Delete Tensor::ShallowCopy", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 33}}, {"node": {"oid": "fc90a9b2fc6e35f601b38d956a0e1e14d90e3046", "messageHeadline": "allocator refactor (#467)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 35, "deletions": 57}}, {"node": {"oid": "0c4fef9ac2993728646e30e2142098667883df40", "messageHeadline": "Jignparm/removemodelcopies (#471)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 111, "deletions": 22}}, {"node": {"oid": "ec8ac04f3041e391abc203ac9c97f6b06e828bcd", "messageHeadline": "Update cast op to support string <-> numeric (#379)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 323, "deletions": 150}}, {"node": {"oid": "f72474c24bea96b2de011fab1d51e28e897d6da1", "messageHeadline": "Updated System requirements in README.md (#466)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "e7c1b774e8956e040e75aee3bf4a7def052b3e05", "messageHeadline": "Move build dependencies like setuptools wheel numpy into docker image\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "892c0653cc771396133defada611fa2501878723", "messageHeadline": "TopK Op: Include support for valid `axis` attribute in implementation\u2026", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 125, "deletions": 53}}, {"node": {"oid": "fdd71574d6b360a3d472118f3d7856dde80ce07a", "messageHeadline": "misc: Fix comment in op_node_proto_helper (#460)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "88949485ff458bcc31b804271cfce8fc0c4aea27", "messageHeadline": "removed MklDnn dependency from C# (#455)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 4, "deletions": 5}}, {"node": {"oid": "e57b5116d6eb83283aab2e33bf08206b9053dc8f", "messageHeadline": "BrainSlice parameter represents the IP.  Update parameter name to match", "author": {"name": "Jesse Benson", "email": "benson.jesse@gmail.com", "user": {"login": "jessebenson"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "aac711ab2f458cf3659c5698f0f44532c905d83a", "messageHeadline": "Implement Sign operator. (#456)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 297, "deletions": 3}}, {"node": {"oid": "360fc32db4b6791dc37c564b974b4d4c87183b6e", "messageHeadline": "compute forward and backward parallel for MLAS and not use_openmp (#457)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 7, "deletions": 10}}, {"node": {"oid": "5d00b8b3756205ecdbe5a1776d39c24e3961c40a", "messageHeadline": "Fix docker gpu test for csharp package cuda 9.1 and 10 (#432)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 12, "deletions": 8}}, {"node": {"oid": "7b37dc6105052efb91f02b621b19e5ec9edacec0", "messageHeadline": "Enable USE_MKLML_FOR_BLAS (#387)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 3, "deletions": 4}}, {"node": {"oid": "db0fde9add109e65ef473fc87a4e2d86e9bd113e", "messageHeadline": "Make USE_MLAS macro conditional on cmake flag for consistency with ot\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "4cdb0cbf6e17223bde554f2853e8c73547344b84", "messageHeadline": "A tiny fix in KernelCreateInfo", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "fb04940ad3c57a154bcde85c96d62824834e2f1e", "messageHeadline": "Initial implementation of NonZero op. (#437)", "author": {"name": "edgchen1", "email": "edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 210, "deletions": 2}}, {"node": {"oid": "7c70d9349a8e176eaff3f9774e9f1da4b659f2d0", "messageHeadline": "Fix a bug in execution_provider.cc", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "657d46fb3c219799e4f07c0a1933f01288098bcd", "messageHeadline": "Output empty shape scalar for empty input. (#451)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 5, "deletions": 8}}, {"node": {"oid": "f20258e9ed584ac3b3e39fbf2c8e72ec98d9840e", "messageHeadline": "Delete dead code", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 29}}, {"node": {"oid": "8a8d1b0cea0df4237e2be2dd4d8d9aa5dbc4a2b5", "messageHeadline": "Fix MacOS shared library build (#447)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 14, "deletions": 5}}, {"node": {"oid": "f14b258a5c9ffb586aa27d3837b6f7778dfaff29", "messageHeadline": "Fix float 16 type support for some CUDA kernels (#436)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 275, "deletions": 15}}, {"node": {"oid": "5866e853c44490a586815df315cea2e8453f02bb", "messageHeadline": "Add dev notes", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 41, "deletions": 0}}, {"node": {"oid": "7cd393d6971bf2d76c925ab68707243485282dd2", "messageHeadline": "Fix 3.7 build; Add cuda version in README (#427)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 4, "deletions": 1}}, {"node": {"oid": "b29c6e48b44c542e9dcc1eb8e91b88e30698e0a6", "messageHeadline": "The files of graph_transformer.h and rewrite_rule.h has been moved. (\u2026", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "405c4bacbc885dd0b81cbd4d83d85cbc3bebd526", "messageHeadline": "Fix a bug in SessionState", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c932ab8e99ff7db7a16801bd12d97bc1b5d0191b", "messageHeadline": "Implement ConstantOfShape (#443)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 428, "deletions": 5}}, {"node": {"oid": "4038db14e28540897d7859658e3558441a094a1e", "messageHeadline": "update trt due to removing reference counting API changes (#444)", "author": {"name": "stevenlix", "email": "Li.Steven@microsoft.com", "user": {"login": "stevenlix"}}, "additions": 3, "deletions": 10}}, {"node": {"oid": "851e291f22c4ca862c86fd7ef84497366d7ca030", "messageHeadline": "Make OpKernelInfo not depend on SessionState. (#442)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 64, "deletions": 24}}, {"node": {"oid": "9faac70daed21aa3a15b7b9575d8ecf5b3e24388", "messageHeadline": "Delete Tensor's copy constructor", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 34}}, {"node": {"oid": "d35409f58ef881e5b7e02533fa9c199e98386dff", "messageHeadline": "Support uint8 datatype for Upsample op in CPU and CUDA providers (#440)", "author": {"name": "Hariharan Seshadri", "email": "shariharan91@gmail.com", "user": {"login": "hariharans29"}}, "additions": 44, "deletions": 0}}, {"node": {"oid": "2062c490337acfc0ea717bf13d1cedf4c5dacfa0", "messageHeadline": "Rashuai/fix dilation (#415)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 147, "deletions": 6}}, {"node": {"oid": "696ab8a194c3237102c713ad3e5cb7003316cd6b", "messageHeadline": "Create a separate component for graph optimization. (#421)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 78, "deletions": 61}}, {"node": {"oid": "737700f94fc2b503c5c87847b9665999efa76b98", "messageHeadline": "fixed the win10 runtime paths to win (#435)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "214c1b88e340272b08ef56d297e97e09543e265b", "messageHeadline": "fix brainslice break", "author": {"name": "souptc", "email": "chenta@microsoft.com", "user": {"login": "chentaMS"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "3b061d60a9f2e19a68359c7c1f9830ef4901045c", "messageHeadline": "Updating new protobuf generated C# file (#430)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 601, "deletions": 55}}, {"node": {"oid": "5cac965471368c716d564e97b71b0ffdcd928941", "messageHeadline": "Copy input tensors (#395)", "author": {"name": "Artem Rudoy", "email": "38088666+artemrd@users.noreply.github.com", "user": {"login": "artemrd"}}, "additions": 21, "deletions": 9}}, {"node": {"oid": "ebfed607411120c19447b56e30bcf653f46a6630", "messageHeadline": "Resync protobuf def", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1512, "deletions": 48}}, {"node": {"oid": "f85cd520c036efead482a0ecf748db35835f2491", "messageHeadline": "Recurse into subgraphs in transformers and session initialization (#368)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 427, "deletions": 268}}, {"node": {"oid": "93bcd9beb6b616f33f723548fa8c180c2bd2239e", "messageHeadline": "Type and Shape inference for QuantizeLinear and DeQuantizeLinear Ops \u2026", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 22, "deletions": 2}}, {"node": {"oid": "60cdb79204abbf786149faecaf9660259e6b5907", "messageHeadline": "Enable tests for EyeLike and enable datatypes present in tests (#424)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 18, "deletions": 13}}, {"node": {"oid": "9f0298261d0bd20a67ecbe29a1ec8091ac6559a0", "messageHeadline": "Fix a build warning in onnxruntime python extension (#416)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 0}}, {"node": {"oid": "d369ff945dbfd9109213cf0ae69d28d3275ba79e", "messageHeadline": "Re-enable python tests (#419)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 55, "deletions": 54}}, {"node": {"oid": "011a784eaa4c61e6f0ae90e616ec815942dab84a", "messageHeadline": "Merge back from rel-0.2.1 (#422)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 10, "deletions": 4}}, {"node": {"oid": "6ae6853519dc4ee85f58ac8de5cb19a9664ecd70", "messageHeadline": "Update test data md5", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "efb72540be4b2a7900178b1e0097779b0baa06db", "messageHeadline": "Separate out constant node index information from ExecutionFrame (#410)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 238, "deletions": 233}}, {"node": {"oid": "fb7be27096ee5492ef9889ce5c37a07946c6fd1a", "messageHeadline": "Update test dataset", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 22}}, {"node": {"oid": "266201de0ad3040b2da8bf6493f0ca44dc5fe8ff", "messageHeadline": "limit thread pool size when running mkldnn model tests", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 1}}, {"node": {"oid": "70985fa803574de98cbf4b86df5386fb79a41d8a", "messageHeadline": "Update BUILD.md", "author": {"name": "DronexAI", "email": "sw6y15@soton.ac.uk", "user": {"login": "sw6y15"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "91ffb980a27cfe9ceb881935dfc30f222544933c", "messageHeadline": "Addl TPN updates (#403)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 487, "deletions": 13}}, {"node": {"oid": "c76725da2d7877a0fc4070598577834ab59886ab", "messageHeadline": "Slice elimination rewrite rule; re-implementation of identity elimina\u2026", "author": {"name": "Konstantinos Karanasos", "email": "kkaranasos@gmail.com", "user": {"login": "kkaranasos"}}, "additions": 185, "deletions": 37}}, {"node": {"oid": "59955da188d32336d29b514d00f53d0dcbe8e36b", "messageHeadline": "Add test case for pre-allocated output", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 39, "deletions": 5}}, {"node": {"oid": "925b3f059cd09a63b1b72ccd1de6d2e4afb26c03", "messageHeadline": "Update cgmanifest.json", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 103, "deletions": 65}}, {"node": {"oid": "ea94cbf14bd4eed59c4e18db6a692ac538652816", "messageHeadline": "Docker containers for CPU and GPU quickstart (#332)", "author": {"name": "Vinitra Swamy", "email": "vinitras@gmail.com", "user": {"login": "vinitra"}}, "additions": 115, "deletions": 0}}, {"node": {"oid": "583631adf5fd2e7d7c92b1f664f809483ae1b19d", "messageHeadline": "Add cgmanifest.json", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 176, "deletions": 0}}, {"node": {"oid": "6fc48c60deac2939837e0f08d57577359115f108", "messageHeadline": "Add win-ci-pipeline-cg.yml", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 135, "deletions": 0}}, {"node": {"oid": "20ef8b43a6df22e6effe71c61e11a0def9406cf0", "messageHeadline": "Change GEMM kernels to natively handle broader range of row counts (#\u2026", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 847, "deletions": 1124}}, {"node": {"oid": "d75bdc519410e6d825eb5624846f67452543fd63", "messageHeadline": "updated C# API doc (#405)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 20, "deletions": 5}}, {"node": {"oid": "09806625cf620606329c2d7be4c97083c57b32dd", "messageHeadline": "Rename OrtInitialize to OrtCreateEnv in preparation for future. (#399)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 25, "deletions": 29}}, {"node": {"oid": "2f73d7abf87d8a29961b3e27b2be26d3c57c3a15", "messageHeadline": "compile with GL/LTCG (#391)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 10, "deletions": 1}}, {"node": {"oid": "f13b9ac4291d7a76b9778bc7a4cb2f8fc8253814", "messageHeadline": "Refine word_conv_embedding (#388)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 47, "deletions": 31}}, {"node": {"oid": "439dbbada9546fa41d8abfe09d94b7129f934fd4", "messageHeadline": "Adds OnnxTransformer to plug onnxruntime in sckit-learn's pipeline (#\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 160, "deletions": 1}}, {"node": {"oid": "7c21c1573269f8bf1619e7cd7dc8847c0cebf396", "messageHeadline": "Jignparm/addcsharptestgpu (#393)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 166, "deletions": 0}}, {"node": {"oid": "68881fadcd5dbb961c23c35f60fa8398a30f027f", "messageHeadline": "Delay load cudart64 for cpu execution", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d4d32708912304df60bf942a0c2b6a484a9fb82d", "messageHeadline": "Update transpose logic for Scan 9 outputs so it's the inverse of the \u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 87, "deletions": 14}}, {"node": {"oid": "b194b7df0d95f0546cfe8c8b9c01ab83c2d2f5bc", "messageHeadline": "Add the ability to use a custom allocator for fetches to avoid unnece\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 181, "deletions": 149}}, {"node": {"oid": "b194d79dfbe1a293c1dd4e1ea25acac9cfbb219a", "messageHeadline": "Third party attribution updates (#398)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 118, "deletions": 9}}, {"node": {"oid": "8f215b44e027513b993b37ef45f1bb78e3af4d36", "messageHeadline": "Refactor InferenceSession::Impl::Load code to remove duplication. (#248)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 77, "deletions": 118}}, {"node": {"oid": "b92bc998619b795b475bdd1e333c476d13a9e2a5", "messageHeadline": "QLinearConv (#370)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 603, "deletions": 3}}, {"node": {"oid": "5ef4c90f1dc3ced8ba844c5d750bd952902f861b", "messageHeadline": "Make the return  namedonnxvalue objects disposable in C# API (#392)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 515, "deletions": 276}}, {"node": {"oid": "571e1e9a6cf15a7dd778ed276513851556aa02e0", "messageHeadline": "Jignparm/updateversion 2.0 (#394)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 10, "deletions": 4}}, {"node": {"oid": "7c0a6f3d9c91525cc8f92d702a58ef652ce7deaa", "messageHeadline": "CI: Enable C# tests", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 36, "deletions": 1}}, {"node": {"oid": "8e03560dbbbbb33db22b8cc0c0f4f78d6e829fea", "messageHeadline": "Fix -e option for runtest-docker.sh (#385)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "d875ab2acdafe415a9eccbef68954fb608a76e9f", "messageHeadline": "C API - Remove reference counting (#344)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 253, "deletions": 651}}, {"node": {"oid": "6349114583f4b5765dae538c7b49759cc6c96f29", "messageHeadline": "Revert \"Rashuai/link with ltcg (#378)\" (#383)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 5}}, {"node": {"oid": "ab730cc58d869004120a10636e0246924dbd8a75", "messageHeadline": "passed the OnnxRuntimeBuildDirectory to the docker for the dotnet lin\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "ccca1e9402eda5477f2f54d24b4e8b23332f6848", "messageHeadline": "Update property file for Nuget Linux package (#369)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "f53cc032db69d1bf5c1af102eb83da7773f0e079", "messageHeadline": "Rashuai/link with ltcg (#378)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "373177ddd3f438698ebc9e84bd9f2932d4ae5a9f", "messageHeadline": "gemm with empty input (#350)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 22, "deletions": 1}}, {"node": {"oid": "f922d427d3f612328dd3e0392e3b4d54435d4cd1", "messageHeadline": "Improve VerifyKernelDef() performance when op has many inputs/outputs\u2026", "author": {"name": "edgchen1", "email": "edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 120, "deletions": 40}}, {"node": {"oid": "644c13050bba29ad448150ee6d477cc9fed42943", "messageHeadline": "Handle negative axes for reduce ops (#365)", "author": {"name": "Bowen Bao", "email": "semisqg@gmail.com", "user": {"login": "BowenBao"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "f94fdad861a66cc06ad2781bc1547e8b70151531", "messageHeadline": "Fixes on the dotnet end-to-end test scripts to get it running on linu\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 44, "deletions": 15}}, {"node": {"oid": "829b2a5e8104477df18e3f18902b1bd2b694d826", "messageHeadline": "Promote TfIdfvectorizer to ONNX ver 9 (#373)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 106, "deletions": 252}}, {"node": {"oid": "89f643f04ba990cb0dac045183dc46396fd1280b", "messageHeadline": "add new types to shape op (#362)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 31, "deletions": 22}}, {"node": {"oid": "61bbf4bfcca32a8bdc96807a223e8d636c401afe", "messageHeadline": "Fix redundant population of output_indices_ in the execution frame an\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 8}}, {"node": {"oid": "82b4ec3d427d181d97fa1856d3eb83590c9345e5", "messageHeadline": "Fix a bug in KernelRegistry::Register function", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "bca8daf76247c649674ef0afc29cb6cc6e055922", "messageHeadline": "Update ONNX. Implement Scan 9 changes (#366)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 298, "deletions": 94}}, {"node": {"oid": "8ea7197b82da0a2d477fea7626f68cc8f04c5ed9", "messageHeadline": "trt (#361)", "author": {"name": "stevenlix", "email": "Li.Steven@microsoft.com", "user": {"login": "stevenlix"}}, "additions": 77, "deletions": 2}}, {"node": {"oid": "904d7c6ec8654bc3b4051455c2f9c9246a4d9889", "messageHeadline": "Add --cuda_version option to enable manually specifying cuda version", "author": {"name": "Harry Summer", "email": "harrysummerxr@gmail.com", "user": {"login": "harrysummer"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "d040b452cb142bc906fbd42ff21ba316551af6de", "messageHeadline": "Expand: add additional supported types. (#364)", "author": {"name": "Bowen Bao", "email": "semisqg@gmail.com", "user": {"login": "BowenBao"}}, "additions": 142, "deletions": 6}}, {"node": {"oid": "ea816615eb4131bd983cf9a9c4510ff9e4fbe43d", "messageHeadline": "remove use_tvm from base script. Put it in yaml configuration (#363)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "647cc2dced212cc3bd7ec276d3b9c5329c5227ba", "messageHeadline": "use gemm to replace matmul + add (#234)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 475, "deletions": 26}}, {"node": {"oid": "8b55596dfefd22793d65ccfecf4cef2d5f1772aa", "messageHeadline": "The CUDA compiler doesn't support gsl::suppress so disable when __NVC\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c87929e949bc545ce8ef737efe283d27f9a8c06b", "messageHeadline": "Use nsync for implementing condition variable", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 331, "deletions": 163}}, {"node": {"oid": "1653ba9fcc638cf3a50f8ee622cfadd7fde11216", "messageHeadline": "Optimizing Upsample op (#352)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 28, "deletions": 21}}, {"node": {"oid": "22337bb64132a049388cc586f0fccd19266ef3fe", "messageHeadline": "fix linaro build (#355)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 6, "deletions": 12}}, {"node": {"oid": "0a21226b096b2bd6c6449819caff593cdb0efcfb", "messageHeadline": "comment out 16-bit float models in C# (#351)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 37, "deletions": 5}}, {"node": {"oid": "6f30bec0406c0c6f387642bb1d2566595718d50f", "messageHeadline": "Implement MLAS convolution+activation fusion (#354)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 512, "deletions": 116}}, {"node": {"oid": "6831fc16ed1df89ea0cad4c8641eeb10ae74988c", "messageHeadline": "Kezhan/kernel registry refine (#346)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 617, "deletions": 604}}, {"node": {"oid": "948cc03490beaff5f01e93a4154c544abd64de36", "messageHeadline": "upgrade onnx", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 4}}, {"node": {"oid": "21713b7a41bc356aa771a6c3fb765b82a1f8195d", "messageHeadline": "Reduce test parallelism for cuda model tests", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "36c62d84b4184b81d7ef1f81428c40c4586b4012", "messageHeadline": "remove ConstantLike OP", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 206}}, {"node": {"oid": "9f3ae4279f7fd0c6af9e6607e8a94964f697a12a", "messageHeadline": "Handle copy to/from non-CPU devices across control flow nodes (#339)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 786, "deletions": 508}}, {"node": {"oid": "c2704b5afb5d983997384cdf3d2fc610f5e3a4cc", "messageHeadline": "cleanup code (#343)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 737, "deletions": 761}}, {"node": {"oid": "b3f0d0b6598f9f8669d263d6d96e1b94989ec7db", "messageHeadline": "added unit test to guard against native API changes (#337)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 47, "deletions": 16}}, {"node": {"oid": "790cda6ea75a75e82bf41ae74a95f355593dfd7b", "messageHeadline": "Fix the issue which causes wrong output. (#342)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 20, "deletions": 6}}, {"node": {"oid": "5d0e024284a84cbeb02846f602914c4324f8e053", "messageHeadline": "Askhade/add quantized matmul (#295)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 479, "deletions": 2}}, {"node": {"oid": "34afa0a5986285d247df748730873b162ec417bb", "messageHeadline": "Delete onnxruntime_exec", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 790}}, {"node": {"oid": "d23f01dcd994decf477adbad9b45ca3669317582", "messageHeadline": "Suppress warnings for gemmlowp", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 22, "deletions": 2}}, {"node": {"oid": "95b8941e9d534bdec2a445d7616e003ba3c953ee", "messageHeadline": "Fix Seg fault when repeats input contain a 0 (#336)", "author": {"name": "Ashwin Kumar", "email": "33531737+ashku-ms@users.noreply.github.com", "user": {"login": "ashku-ms"}}, "additions": 30, "deletions": 1}}, {"node": {"oid": "f678f58750c32f5b92aa8a44957ef314fa04bba2", "messageHeadline": "Revert to ignoring optional subgraph inputs (#306)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 43, "deletions": 8}}, {"node": {"oid": "6225d5fe1e19653ff6863bbfb314c1c9a1c9aa9b", "messageHeadline": "Update test data (#334)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "492d9fd6cc963142f8e61e1b0599ab81705ddaf6", "messageHeadline": "Use Eigen ThreadPool in OnnxRuntime (#323)", "author": {"name": "Ashwin Kumar", "email": "33531737+ashku-ms@users.noreply.github.com", "user": {"login": "ashku-ms"}}, "additions": 283, "deletions": 193}}, {"node": {"oid": "139abda393d8410e1c35469adfcc7d13d109a997", "messageHeadline": "convinteger implementation based on gemmlowp (#294)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 252, "deletions": 61}}, {"node": {"oid": "835b511fa81119ff5cf350ab93e2a1f709ed1d8a", "messageHeadline": "cuda fix to unblock the tf model tests (#333)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 32, "deletions": 5}}, {"node": {"oid": "797787174013bae96f9bfafcddd9854c911efd72", "messageHeadline": "Split build pipeline", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 231, "deletions": 61}}, {"node": {"oid": "7e3923b9b305407a736ae2ddb901301ade8171b9", "messageHeadline": "Fix for non-wide characters in strings for linux - for c#-native inte\u2026", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 38, "deletions": 30}}, {"node": {"oid": "779123cf556f9a663d49daefab623322378eefc2", "messageHeadline": "Upsample opset 9 cuda implementation (#330)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 56, "deletions": 39}}, {"node": {"oid": "0efc48a11aeaea633dbd7c977e7a037318e2a40c", "messageHeadline": "Install dotnet sdk on linux ci (#320)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 9, "deletions": 0}}, {"node": {"oid": "677918cd9a7c1debeaf4e6a70efe4979905e4376", "messageHeadline": "Added generation of C# project properties file containing actual buil\u2026", "author": {"name": "Edward Chen", "email": "edch@microsoft.com", "user": {"login": "edgchen1"}}, "additions": 30, "deletions": 16}}, {"node": {"oid": "c955cd827865b0e6a8853a5d0010f0d0c5228bf2", "messageHeadline": "changed csharp runtime folder name to win from win10", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 5, "deletions": 5}}, {"node": {"oid": "25c1e68988f21d28608df2c2c656666c567e420d", "messageHeadline": "Fix: roi_pool operator implementation error about FLT_MIN", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ef5679949a98c56d766e1fe370f31299c2e75f5d", "messageHeadline": "Fix a c# build issue when mkldnn is not enabled (#321)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "3b83f062fcddd27e145d58ae8c9b2038fee2bc78", "messageHeadline": "remove delayload from mkldnn (#276)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "260639c3276719ebfa9ffe05ee04a44c262ed182", "messageHeadline": "Add missing EXCLUDE_FROM_ALL keyword to nsync submodule", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "2d067ec65c3e40ba83b928785628e0906bcb6a02", "messageHeadline": "Update with link for C# GPU Nuget package", "author": {"name": "Faith Xu", "email": "faxu@microsoft.com", "user": {"login": "faxu"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4735bb1ccb044f603970f1b120b206b5b3af8252", "messageHeadline": "Add onnx protobuf format that supports large model (#313)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 53, "deletions": 38}}, {"node": {"oid": "05b9440fceabc8b95b52173ba67330b4540ac5c1", "messageHeadline": "mkldnn:Conv weight optimization (#256)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 77, "deletions": 21}}, {"node": {"oid": "8c40313e281071d0adab6cd226c6366c261d1cc6", "messageHeadline": "Update documentation to reflect the latest changes (#311)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 51, "deletions": 51}}, {"node": {"oid": "7641ee9a2bfe9f7fe6ee988e062a95175827f5cc", "messageHeadline": "suppress a warning.", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "d22429c5b27f0b6e6fa209f8c1804a276007c103", "messageHeadline": "Update compare_mlvalue for tests (#290)", "author": {"name": "Bowen Bao", "email": "semisqg@gmail.com", "user": {"login": "BowenBao"}}, "additions": 11, "deletions": 4}}, {"node": {"oid": "02962ce9d866b602d8b57cf7fef69cf12d7ebf65", "messageHeadline": "Update ABI.md (#299)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 47, "deletions": 46}}, {"node": {"oid": "02852a0881948e6f8298d6b94ed8aaaf07964572", "messageHeadline": "Remove OperatorParser tool (#279)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 0, "deletions": 80}}, {"node": {"oid": "223773d278e109f1938788cd34ba628adb009514", "messageHeadline": "Implement ROI Align for object detection. (#308)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 712, "deletions": 25}}, {"node": {"oid": "6b3044ddd35310bac13d52357f1ebf39de89e38a", "messageHeadline": "Update AddingExecutionProvider.md", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 35, "deletions": 10}}, {"node": {"oid": "fa0ea9a273a1a8a67cb8c9c73eca09954a848e01", "messageHeadline": "implement dynamic slice cuda (#286)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 73, "deletions": 52}}, {"node": {"oid": "98a92547bffdee2ace73c2aa8773eed018fdc283", "messageHeadline": "Ryanunderhill/c api 8 (#297)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 154, "deletions": 203}}, {"node": {"oid": "751eb60819ca3d47f44c8554a11326bdaf95a1f9", "messageHeadline": "Fix a build issue in onnxruntime_unittests.cmake (#307)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d342147255d53348b371593fc5d202343a6f2aa9", "messageHeadline": "Add test for truncated sequence inference (#99)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 294, "deletions": 18}}, {"node": {"oid": "34bcc92554fd6e0581fc88d94a8acb8a8d5f25a6", "messageHeadline": "Added test data URL and checksum arguments to build.py. (#302)", "author": {"name": "edgchen1", "email": "edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 115, "deletions": 91}}, {"node": {"oid": "55433abecbd7e97efc9fa7b7280ec47a84370a12", "messageHeadline": "Enable build_shared_lib if build_csharp", "author": {"name": "Yufeng Li", "email": "roli@microsoft.com", "user": {"login": "yufenglee"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "84be07114bb00c6d8a93200089a6f1ee0f582824", "messageHeadline": "Disable nsync test", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 0}}, {"node": {"oid": "22b5bfdf35125e2fee296206ab430eeb8da05da5", "messageHeadline": "update label for build badges (#301)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "e318c7317bb949de9ee4ad6fc7a92bd49e1f81d5", "messageHeadline": "update", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "8cfe8d33a3da698e27fe1d62ad36c7b3b14cdbe6", "messageHeadline": "Add nsync (#292)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 128, "deletions": 14}}, {"node": {"oid": "57421504b1a37f9b6fa97896867c185545e1cb29", "messageHeadline": "Add tensorflow license (#296)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "31bbb4598e033318513b1604f245673d3839f001", "messageHeadline": "Enable tvm in CI builds. (#285)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "386b60e35ab8ecff16961cde9f195896e0e0bdd0", "messageHeadline": "update gemmlowp which carries a windows build break fix. (#291)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4929ddde4317c67bec6465ff8966138a19c287b3", "messageHeadline": "update tvm submodule (#287)", "author": {"name": "Yang Chen", "email": "40417152+yangchen-MS@users.noreply.github.com", "user": {"login": "yangchen-MS"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "059cfdc13a0f95795b7ca3e99d25fb6613167b08", "messageHeadline": "Opset 9 Scan implementation (#274)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 1992, "deletions": 1012}}, {"node": {"oid": "5e113661a9e1f60f47aeffbd598105e91e1d8e49", "messageHeadline": "Build system upgrades (#281)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 444, "deletions": 342}}, {"node": {"oid": "85ec13f58df7ecec3c81b46b11ec90fa8902f364", "messageHeadline": "fix tvm break (#282)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 22, "deletions": 16}}, {"node": {"oid": "ec2cf59baaee522da4478f826f2a8caeacd14ff2", "messageHeadline": "Enable building python37 packages (#283)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 6, "deletions": 8}}, {"node": {"oid": "de383d93be51df96b01de217bb9526546cf5bbe0", "messageHeadline": "Fix inconsistency in enum names in the C API (#277)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 10, "deletions": 10}}, {"node": {"oid": "251d4bbb230ffadb3a7931019b6d0279caaa4b77", "messageHeadline": "adjust the quantization ops' definitions (#278)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 35, "deletions": 8}}, {"node": {"oid": "8a7b6761dec568d01dad7677875b9f07b018484b", "messageHeadline": "changed nativeLib name from onnxruntime.dll to onnxruntime, for cross\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "75934af8967f219eb7b67d32fda45e5ae035a782", "messageHeadline": "have Im2ColNd support all types and allow customized padding value. (\u2026", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 114, "deletions": 108}}, {"node": {"oid": "058803086d6a41d047ce84e53ea3578748fdfbc9", "messageHeadline": "Implement N-gram (#180)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1314, "deletions": 0}}, {"node": {"oid": "8fba324678dd4f1267043f669ddefba0806c06fe", "messageHeadline": "Hyperbolic inv ops (#272)", "author": {"name": "ashku-ms", "email": "33531737+ashku-ms@users.noreply.github.com", "user": {"login": "ashku-ms"}}, "additions": 118, "deletions": 6}}, {"node": {"oid": "d0fa974976bc6e9fd104349a6901e708d009fd1d", "messageHeadline": "interface change to code-generated kernels (#192)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 727, "deletions": 185}}, {"node": {"oid": "5b6f1823af859e62fac8afdf14656b7f9b5df7f7", "messageHeadline": "Use CUDA libraries found by CMake. (#190)", "author": {"name": "xkszltl", "email": "xkszltl@users.noreply.github.com", "user": {"login": "xkszltl"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "fc76076e29313fa0e3daaf29f2b829e079ac7016", "messageHeadline": "Rashuai/dynamic slice refactored (#264)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 320, "deletions": 64}}, {"node": {"oid": "bd2ace7619821ff63ec7e8c659f802e2616a40ec", "messageHeadline": "Tweak linker flags to minimize the binary size. (#270)", "author": {"name": "Yuan Yu", "email": "yuanbyu@gmail.com", "user": {"login": "yuanbyu"}}, "additions": 9, "deletions": 1}}, {"node": {"oid": "e7e90c0322e9393cbd41db29760a979907321f1e", "messageHeadline": "Implement OneHot op.  (#213)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 360, "deletions": 4}}, {"node": {"oid": "bddeb3d001c91a37620df468d9efef75906c65ac", "messageHeadline": "Fix \"invalid escape\" issue when eigen_SOURCE_PATH contains \"\\\". (#258)", "author": {"name": "xkszltl", "email": "xkszltl@users.noreply.github.com", "user": {"login": "xkszltl"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "eb67eadbc64d7cebd6f7602351882697ee8d4b2f", "messageHeadline": "add initializer for sub-graph. (#269)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 8, "deletions": 0}}, {"node": {"oid": "8f3a492682941559fe3496865b1fb889883610fc", "messageHeadline": "Remove static linked cuda libs (#239)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "e2746513ab2536222fe957f1d5e4a95de3fcd9ba", "messageHeadline": "Avoid to run profiling code completely if there is no need (#245)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 105, "deletions": 66}}, {"node": {"oid": "1e9be01a49a96d43b0f17f18d9a6dfad01227274", "messageHeadline": "Fix a bug in Conv+Activation fusion (#236)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 29, "deletions": 7}}, {"node": {"oid": "4f49a4ab1bac4b6c334741c999ffbfccbd66fa23", "messageHeadline": "Maxpool_With_Mask (#267)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 333, "deletions": 19}}, {"node": {"oid": "b508835e7a250d32e3cab79c67cf01b99a9ab356", "messageHeadline": "Update CUDA Gemm to opset 9 (#266)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 8, "deletions": 7}}, {"node": {"oid": "126c1fd3df628a99dec6c0c29b53345f8c422a63", "messageHeadline": "do replacement based on node arg usage, instead of provider type (#263)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 47, "deletions": 8}}, {"node": {"oid": "6a090985fbbf74c0fa3dda19537f3c5160a506b5", "messageHeadline": "More C API changes (#259)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 154, "deletions": 172}}, {"node": {"oid": "29d03ffb08e3cace78e10101cda2257567fa7f49", "messageHeadline": "Add end-to-end test to run on the nuget package (#252)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 137, "deletions": 4}}, {"node": {"oid": "8380e5640951b730a8dcaa017567be43103c89e3", "messageHeadline": "Optimizing Gather op. (#262)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 12, "deletions": 12}}, {"node": {"oid": "928acc32f1bae56bf39ac2d45d29e77cda6654b8", "messageHeadline": "Minor formatting proposals", "author": {"name": "Dar\u00edo Here\u00f1\u00fa", "email": "magallania@gmail.com", "user": {"login": "kant"}}, "additions": 8, "deletions": 8}}, {"node": {"oid": "698ebf13a170e373a9c1420d806474dc104cb252", "messageHeadline": "Add a missing header (#243)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "79c623bece50782f0fdcad0eb335b69a09e10117", "messageHeadline": "Add more test models", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 61, "deletions": 20}}, {"node": {"oid": "f3df7e5d3238b9dd2f2861edb14d263824e14603", "messageHeadline": "nullptr", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 12, "deletions": 5}}, {"node": {"oid": "b93eba17c7259ac572af3e56166ad3adf417fb58", "messageHeadline": "mkldnn: relu, sum and batch norm (#238)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 796, "deletions": 4}}, {"node": {"oid": "3875511f9e6bc2c7d16ca4d78e4af001d5caea07", "messageHeadline": "Fix inefficiencies in the mkldnn kernels. Some of these were (unfortu\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 16, "deletions": 17}}, {"node": {"oid": "7af1887b330fc8b768c638488441ed5e956c1f2c", "messageHeadline": "Introduce basic BFloat16 runtime support (#235)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 324, "deletions": 76}}, {"node": {"oid": "4e74ffba9122695555754e5ec14a3faf094ce24f", "messageHeadline": "Add word conv embedding custom op (#229)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 455, "deletions": 2}}, {"node": {"oid": "a37887cfa1974ac5fdd986ed15eb55b12d1b10b4", "messageHeadline": "More intuitive ordering to the API functions (#233)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 226, "deletions": 235}}, {"node": {"oid": "c453b48b71ae8a858d466b2ed45827b2cc26d591", "messageHeadline": "update kernel memory type interface (#225)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 59, "deletions": 37}}, {"node": {"oid": "a43382e390dd9244cc5e46f308ff19d9e4aa311c", "messageHeadline": "Jignparm/csharp gpu (#221)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 73, "deletions": 68}}, {"node": {"oid": "a19b6243025ff40fc1c85fc12f0a3028438e2853", "messageHeadline": "MaxUnpool Operator - CPU Implementation (#177)", "author": {"name": "ashku-ms", "email": "33531737+ashku-ms@users.noreply.github.com", "user": {"login": "ashku-ms"}}, "additions": 652, "deletions": 4}}, {"node": {"oid": "eb867be33190e09b9fdad1fa42a434d24900930a", "messageHeadline": "update mkldnn to 0.17.2 (#231)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "abce6041c16b97545becacb1df5055bacfa55e99", "messageHeadline": "Print hex value for float compare when test failed (#228)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 10, "deletions": 4}}, {"node": {"oid": "0dca08023894857fcc88c4a53ffb9a92b78287c5", "messageHeadline": "remove useless internal schema file (#226)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 11, "deletions": 35}}, {"node": {"oid": "1d95c939874d706d591bf1dbb8976792470f5def", "messageHeadline": "[optimization] avoid vector copy and reduce allocation. (#203)", "author": {"name": "Weixian", "email": "ideazwx@gmail.com", "user": {"login": "zwxxx"}}, "additions": 28, "deletions": 7}}, {"node": {"oid": "255ee39af6c68b64dcff27fa82c556a0d7fc5cc4", "messageHeadline": "Fix memory leak by improper handling of std::string typed (#227)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 15, "deletions": 16}}, {"node": {"oid": "e97caa77875ea69a1345f68ccb002ac0dbd9ef2c", "messageHeadline": "change mkldnn so path (#210)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 24}}, {"node": {"oid": "84231ba0033ff690773ed46b8dae6f62c8e3549a", "messageHeadline": "support hyperbolic ops (#223)", "author": {"name": "ashku-ms", "email": "33531737+ashku-ms@users.noreply.github.com", "user": {"login": "ashku-ms"}}, "additions": 73, "deletions": 25}}, {"node": {"oid": "e63572c1f3a2922388eca309991008a519435f6c", "messageHeadline": "Updated ArrayFeatureExtractor op to retain old output shape behavior.", "author": {"name": "Edward Chen", "email": "edch@microsoft.com", "user": {"login": "edgchen1"}}, "additions": 7, "deletions": 3}}, {"node": {"oid": "94f8f2b05c2ae1a2bdb427919f49d7e2ad263a8c", "messageHeadline": "placeholder for internal contrib ops (#219)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 80, "deletions": 44}}, {"node": {"oid": "b9cc134576f5090fc8e0029ff894f471e86c068d", "messageHeadline": "Make sure tensor sizes are 64-byte aligned (#222)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 3, "deletions": 3}}, {"node": {"oid": "4d010fb1eabaaa38fcd0939043c4a9e54bca1f78", "messageHeadline": "Add null check before calling node.op_->Deprecated(). (#211)", "author": {"name": "Bowen Bao", "email": "semisqg@gmail.com", "user": {"login": "BowenBao"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ac3a081ec56bac62adc869b704416bb2f2f209d2", "messageHeadline": "Enable release build in Windows CI pipelines (#220)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "0248390e4d8c03afe29d0a07e2d0558fed7cb40f", "messageHeadline": "Add support for checking for F16C support (https://en.wikipedia.org/w\u2026", "author": {"name": "Jesse Benson", "email": "jessebenson@users.noreply.github.com", "user": {"login": "jessebenson"}}, "additions": 3, "deletions": 0}}, {"node": {"oid": "ab350fa4c7f3c98a5792a48042673eaaeba99a88", "messageHeadline": "Re-structure the inference session initialization to (#217)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 130, "deletions": 80}}, {"node": {"oid": "334e329642d6d565755892cefff3b815bb866423", "messageHeadline": "Increment/decrement UseCount for outputs so that we don't prematurely\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "39f47f86eecdd74f7cce831ff42eee49a932964b", "messageHeadline": "Adding the include folder for the C Windows pkg. (#198)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 15, "deletions": 3}}, {"node": {"oid": "beb326f00ed7a3e7ae31c5b7aa7cda8e091beb1e", "messageHeadline": "Simplify logic around creating relationship between nodes for implici\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 16}}, {"node": {"oid": "dc8b37f4c491872f75d4fe1963525f98fdb9974b", "messageHeadline": "update onnx (#209)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 21, "deletions": 21}}, {"node": {"oid": "37b74c771aff134fe6af3715ab687f789b9c2c4e", "messageHeadline": "add gemmlowp as submodule. (#206)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "c0ec7d56c4a3ea5ec91ec7f1494535c9ec61cb76", "messageHeadline": "Clean up garbage files (#208)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 65}}, {"node": {"oid": "d4131a31d90125e3d6ff3f710e0a86248020d15b", "messageHeadline": "Disable csharp pretrained tests temporarily (#207)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0aa1b54aaa85294ee6315d05cd33804e034baba5", "messageHeadline": "Updated build.py to support relative CMake/CTest paths and did some m\u2026", "author": {"name": "edgchen1", "email": "edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 37, "deletions": 28}}, {"node": {"oid": "773114a4f125d0cb515526351f5e3defffbca034", "messageHeadline": "More C header naming changes (#202)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 410, "deletions": 415}}, {"node": {"oid": "0287019c1df9c242508534831b8401984c393f2d", "messageHeadline": "C# Gpu : Minor updates to exception message (#201)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 20, "deletions": 14}}, {"node": {"oid": "71c56b6d7cc282a21168c74c3b7afd94594a1465", "messageHeadline": "Fix array feature extractor out of bounds access issue (#194)", "author": {"name": "edgchen1", "email": "edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 109, "deletions": 22}}, {"node": {"oid": "8c5d10555784001d893532b62d850916b3c9ad0d", "messageHeadline": "fix typo", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "b0f27ba0a7cebf54b0c16237b5b81e1191ce8480", "messageHeadline": "Allow using MKLML header/libs when use_mklml is specified (#178)", "author": {"name": "KeDengMS", "email": "kedeng@microsoft.com", "user": null}, "additions": 116, "deletions": 60}}, {"node": {"oid": "d0544a80824de2fd3d18cc60dd39f99f2c781d9d", "messageHeadline": "Rashuai/gathernd op (#170)", "author": {"name": "Randy", "email": "45701928+RandyShuai@users.noreply.github.com", "user": {"login": "RandyShuai"}}, "additions": 387, "deletions": 0}}, {"node": {"oid": "82d04412a0bdfe713a928ce970cfa84b91bf5ffc", "messageHeadline": "Kezhan/partition logic update (#164)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 46, "deletions": 20}}, {"node": {"oid": "5a8acd7da8b28a8380829ba859d9d61a02af69b3", "messageHeadline": "directly updating dst arg with src arg when adding an edge. no need u\u2026", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "5fd902413917a1201d5a46eac4d05336d669f9a6", "messageHeadline": "[WIP] Initial checking for CSharp GPU support (#176)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 107, "deletions": 17}}, {"node": {"oid": "b418adff42824a45ba96e9af272b986f995d3f2b", "messageHeadline": "Disable test data downloading by default (#193)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "383315a7e069845eb9f1f23e550f525122255f8b", "messageHeadline": "Upgrade Eigen to 3.3.7 (#185)", "author": {"name": "xkszltl", "email": "xkszltl@users.noreply.github.com", "user": {"login": "xkszltl"}}, "additions": 11, "deletions": 3}}, {"node": {"oid": "16b3a9fd6353314ccb9b882d6d227672bce7b34d", "messageHeadline": "onnxruntime_CUDA_HOME should only be used when available.", "author": {"name": "Tongliang Liao", "email": "xkszltl@gmail.com", "user": {"login": "xkszltl"}}, "additions": 4, "deletions": 2}}, {"node": {"oid": "6cd935e184d09ed22c3729e6d0e12179adcde044", "messageHeadline": "Duli/fusedconv (#183)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 6, "deletions": 0}}, {"node": {"oid": "7560538c94840cfce03419dec3cecef461d5233c", "messageHeadline": "Revert \"Adding missing attribute *alpha* to FusedConv schema.\" (#182)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 0, "deletions": 62}}, {"node": {"oid": "3f4e62a49cdd416ac1725e30be46f81c916ce7b8", "messageHeadline": "Adding missing attribute *alpha* to FusedConv schema.", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 62, "deletions": 0}}, {"node": {"oid": "55a05276407ae40ce5f91a517ec55c7aa00ac747", "messageHeadline": "Fix a warning in gcc 7", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "11b369a8641d9dc8989739f3c530a8cf330d7f8b", "messageHeadline": "Abbreviate ONNXRuntime as Ort in all of our public APIs (#175)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 2764, "deletions": 2767}}, {"node": {"oid": "47551da99451a1f20b34bf83e06c6ccbd638fe13", "messageHeadline": "Optimize Tanh/Sigmoid activations (#162)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 1261, "deletions": 8}}, {"node": {"oid": "afd831ae9ec49d9cf3e6dc2a9f308a80eb35f8af", "messageHeadline": "Remove unnecessary const", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 11, "deletions": 11}}, {"node": {"oid": "d6abc381826ed0f50c41ee7a9acfcf2025abf9ae", "messageHeadline": "MurmurHash3 operator (#174)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 351, "deletions": 0}}, {"node": {"oid": "fb021b9002061c5b4aa993a98d295074b7b78be1", "messageHeadline": "delete azure-pipelines.yml", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 93}}, {"node": {"oid": "2ffaa8a185c7e2ef141c805ca1ede36b9c75121d", "messageHeadline": "optimize unordered_map (#166)", "author": {"name": "Weixian", "email": "ideazwx@gmail.com", "user": {"login": "zwxxx"}}, "additions": 12, "deletions": 11}}, {"node": {"oid": "0aaaf4663d90eef513be804109940e4959b5a26c", "messageHeadline": "Update data download script (#171)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 57, "deletions": 77}}, {"node": {"oid": "d419170c7b4f7070ac194ff99d4e9b11a2097673", "messageHeadline": "Rename ONNXRUNTIME_API_STATUSCALL to ONNXRUNTIME_API_CALL since it's \u2026", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 54, "deletions": 54}}, {"node": {"oid": "c5a0119d423b3143f2b865fde135b4a13e629b8e", "messageHeadline": "Added Environment::IsInitialized() and added check to InferenceSessio\u2026", "author": {"name": "edgchen1", "email": "edgchen1@users.noreply.github.com", "user": {"login": "edgchen1"}}, "additions": 62, "deletions": 3}}, {"node": {"oid": "34175826dfd3043ec1ef28bab28d6441e9ab77dd", "messageHeadline": "Ryanunderhill/pad fix (#151)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 84, "deletions": 1}}, {"node": {"oid": "1d32aa98f83d77faa01b38083b0d50b515c0f20c", "messageHeadline": "add int16/uint16 in quantization for cntk quantized speech model. (#163)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 13, "deletions": 11}}, {"node": {"oid": "30ab01e7bead5ba1a28edcf5295c07813fac104b", "messageHeadline": "Ryanunderhill/concat neg axis (#152)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 51, "deletions": 5}}, {"node": {"oid": "cfd3b4c6060cbcd7e107fbb5bfb967b09a7486e4", "messageHeadline": "remove extra const (#159)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ed98d3d6535d5c63ee3ff9839f5ad76438e932f5", "messageHeadline": "Add some nullptr checks (#160)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 10, "deletions": 7}}, {"node": {"oid": "618cc51754306ffc41fa0abd027eef2f429ed361", "messageHeadline": "Update onnx_backend_test_series.py (#146)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 90, "deletions": 53}}, {"node": {"oid": "ec71f363f7f9404cf0a82343e5c9970da42d9d6a", "messageHeadline": "Fix a bug in parallel executor and enable test for it in CI (#157)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 25, "deletions": 18}}, {"node": {"oid": "1ed72251d098e9f91c48f115ed67cbc893c19a0e", "messageHeadline": "A tiny change to onnxruntime::Model (#158)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "9baaf5e9566f11b46a8468a5b78ddb89f89f067a", "messageHeadline": "Kezhan/change edge api to use index (#138)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 161, "deletions": 90}}, {"node": {"oid": "2e79597531c44e3afbcad31a3c6234e35cd32459", "messageHeadline": "Revert \"mkldnn Relu/Sum/BatchNormalization kernels\" (#147)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 4, "deletions": 818}}, {"node": {"oid": "2230e5b431da99a995f5fbce8b6ac3876ea808b9", "messageHeadline": "Conv+Activation fusion for CPU (#105)", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 537, "deletions": 155}}, {"node": {"oid": "7f0e5269bd7f04219508a35934d184a64261d63f", "messageHeadline": "Removing from broken tests since this failure cannot be reproduced an\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "f189b76f9ac95338e0cc9c8ef7de9054177ae3b4", "messageHeadline": "Some small edits and renaming. (#153)", "author": {"name": "Yuan Yu", "email": "yuanbyu@gmail.com", "user": {"login": "yuanbyu"}}, "additions": 89, "deletions": 73}}, {"node": {"oid": "0ae9354db1e038df38bd291ea51521c964905d23", "messageHeadline": "Delete unused files in onnx_test_runner (#149)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 191}}, {"node": {"oid": "396e95bc4b8341309a9ba28c2e6e0dc69d9ee421", "messageHeadline": "Reduce whitespace at top of readme for easier reading (#148)", "author": {"name": "Prasanth Pulavarthi", "email": "prasantp@microsoft.com", "user": {"login": "prasanthpul"}}, "additions": 1, "deletions": 4}}, {"node": {"oid": "60c6b9dfd090a89097d441e02b0140155535e651", "messageHeadline": "Kezhan/separate op def from kernels (#143)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 61, "deletions": 44}}, {"node": {"oid": "e2a90ab747b3ea1d250f257aabda38325bf3c2f5", "messageHeadline": "Add unique identifier in function subgraph node (#129)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1beeccb89df7787495550f95cca30f417fbe0647", "messageHeadline": "Enable data cache (#144)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 38, "deletions": 13}}, {"node": {"oid": "8510ccdfe7d62ff0a3a9edf51d683c3b490538cc", "messageHeadline": "Update performance_runner.cc (#142)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b054646ddd9cccaee90d43dfae7620fa2d3f2c2b", "messageHeadline": "Askhade/implement erf (#137)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 38, "deletions": 1}}, {"node": {"oid": "7d79bfef719b4e1afa71eb1954137c814109427e", "messageHeadline": "Move isnan out of contrib_ops and add float16 support for it as per t\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 101, "deletions": 91}}, {"node": {"oid": "9bf78e1f3e6f5335242d7955c70cda5a687bc862", "messageHeadline": "Update BUILD.md (#140)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 19, "deletions": 17}}, {"node": {"oid": "7fb44b9d9ed674fd9f83eeafcae6a08b225a057c", "messageHeadline": "Add missing #include (#139)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c5a691d67d61a9d18f457f341a68c33e5ced825c", "messageHeadline": "update onnx to latest commit (#132)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 9, "deletions": 2}}, {"node": {"oid": "69a5ff3300662e3aecda60781589d7082d433bc8", "messageHeadline": "Upsample opset 9 support for CPU provider (#135)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 102, "deletions": 32}}, {"node": {"oid": "d71376010b5a7d511550abcd6d953c11eeced8a7", "messageHeadline": "Implement Dynamic Range Operator with shape inference", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 389, "deletions": 0}}, {"node": {"oid": "bdd5e5854682a7aefc77e1e94f373ef524bc431f", "messageHeadline": "MKLDNN Sum and Batch Normalizaton kernels (#115)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 573, "deletions": 4}}, {"node": {"oid": "6bfb19518417035c0020a44cce343ed29f70a459", "messageHeadline": "Jignparm/csharp test difftypes (#126)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 4553, "deletions": 15}}, {"node": {"oid": "50ed36290c3cc677b20e11d630f1a1f3fb8387af", "messageHeadline": "Use windows specific C-runtime based solution for lowercasing/upperca\u2026", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 97, "deletions": 36}}, {"node": {"oid": "ed5abc8b94ac4fb34646e08dcf312cd2010bafaf", "messageHeadline": "mkldnn conv 1d, 3d support. (#130)", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 103, "deletions": 78}}, {"node": {"oid": "a09a3d3aa5ec83c0f39c648a45a66aa49bbe06f3", "messageHeadline": "Add build targets and props file for legacy csharp project support (#\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 48, "deletions": 0}}, {"node": {"oid": "19c630b87f1f74940f2b50c03bbb4ab7030a58ba", "messageHeadline": "[TVM] Bugfix. (#117)", "author": {"name": "Siva", "email": "sivar.b@huawei.com", "user": {"login": "srkreddy1238"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "23e122f3048d61e39c01ecfb22c2d88b0466d323", "messageHeadline": "Run backward gru in main thread for bidirection (#118)", "author": {"name": "Yufeng Li", "email": "liyufeng1987@gmail.com", "user": {"login": "yufenglee"}}, "additions": 9, "deletions": 16}}, {"node": {"oid": "6169f3c619fdd432617426d962adb4ac2ab963ab", "messageHeadline": "First Draft EyeLike CPU OP9 (#121)", "author": {"name": "Ashwini Khade", "email": "askhade@microsoft.com", "user": {"login": "askhade"}}, "additions": 180, "deletions": 0}}, {"node": {"oid": "5cec70df4b71a5c7acc76c512b6b44f43158a784", "messageHeadline": "added the redirected URL to nuget package icon (#125)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "c5cbdd5a55a28e84d2d74a74267a288828ceafaa", "messageHeadline": "Miscellaneous fixes (#123)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "a53d617b7ab713d18c01b01daf79e7e5e2c20a60", "messageHeadline": "added the icon and logo resources (#124)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "9b9cb32f882f91ca94d53b39202d70ad86c540f7", "messageHeadline": "Jignparm/csharp test difftypes (#122)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 249, "deletions": 226}}, {"node": {"oid": "830c341c19e4ba35c813adc8c7d1c3de7529f15e", "messageHeadline": "remove graph editor since it's not designed as expected to restrict g\u2026", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 7, "deletions": 60}}, {"node": {"oid": "194d0a98d16527e98228864710fc2e323688cfe2", "messageHeadline": "Add domain and version check for ops fusion. (#108)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 48, "deletions": 6}}, {"node": {"oid": "3c7c1068e73bb2bfab47475aa2ebbf6630b9ea06", "messageHeadline": "refactor threading (#110)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 256, "deletions": 326}}, {"node": {"oid": "6d80253502686208190c81e21053caeb78f797c9", "messageHeadline": "mkldnn activations.relu (#86)", "author": {"name": "Sreekanth Yalachigere", "email": "17345104+sreekanth-yalachigere@users.noreply.github.com", "user": {"login": "sreekanth-yalachigere"}}, "additions": 246, "deletions": 1}}, {"node": {"oid": "c52636e187eefaea0f9376249f9f7ba6ec4beb91", "messageHeadline": "Implement Tokenizer op (#31)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 1440, "deletions": 0}}, {"node": {"oid": "a68f5ccfd91cf053dcb102e14599c4897fa92b1e", "messageHeadline": "Upgrade gpu build to CUDA 10 + cudnn 7.3 (#112)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 117, "deletions": 14}}, {"node": {"oid": "4801e671043818530e2d4fc3c57729081e9f4537", "messageHeadline": "Hecli/cuda10 upgrade (#111)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 21, "deletions": 0}}, {"node": {"oid": "eea7618d53a9db3919b8004278b83f928c87e972", "messageHeadline": "Ryanunderhill/c api (#85)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 597, "deletions": 709}}, {"node": {"oid": "dd1b16dd585a4f3f1409958c06acd88cec68e473", "messageHeadline": "add brainslice change in common folder (#109)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 8, "deletions": 2}}, {"node": {"oid": "3dcf344f093089d1b481faf40bf0a38f2cc83a91", "messageHeadline": "switch build agents to the CUDA 10 pool (#106)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "05739524992eb46919a506a8a175c35959c3208f", "messageHeadline": "Update the documentation, run all examples during the generation of t\u2026", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 152, "deletions": 16}}, {"node": {"oid": "eeb862ab1afde183453ef2eba1884e4c925aeadf", "messageHeadline": "Update Dockerfile for ARM build and the build instruction (#93)", "author": {"name": "Klein Hu", "email": "bohu@microsoft.com", "user": {"login": "NonStatic2014"}}, "additions": 8, "deletions": 5}}, {"node": {"oid": "c8ae91251ea44aaae7ae74ff1c4925536f57d3a2", "messageHeadline": "Eliminate the confusing double negative (#92)", "author": {"name": "R. G. Esteves", "email": "rgesteves@gmail.com", "user": {"login": "rgesteve"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "358be0653c8bc25e2531152f3186b0035b427c27", "messageHeadline": "Delete logo as per marketing requirements (#98)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "996d6ea4cda141bd0217bb694d40e35fea851dde", "messageHeadline": "Revert Softmax optimizations using openmp. (#97)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 8, "deletions": 50}}, {"node": {"oid": "7f0e947f969ef6a7134fcd9cb81c22645b5ee1a3", "messageHeadline": "Add the split build badges (#96)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 3, "deletions": 1}}, {"node": {"oid": "3b08c6665aaedcbbeb1a76e8dad742682fac9c33", "messageHeadline": "Split the CI pipelines (#94)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 96, "deletions": 0}}, {"node": {"oid": "fbb23a9ed004fd6b4b459492d08cbffcfbc81640", "messageHeadline": "Implement StringNormalizer (#69)", "author": {"name": "Dmitri Smirnov", "email": "yuslepukhin@users.noreply.github.com", "user": {"login": "yuslepukhin"}}, "additions": 540, "deletions": 5}}, {"node": {"oid": "005f9dca9650c048fc960f91c681a32dac1ebcfe", "messageHeadline": "Kezhan/renaming graph_base.h to graph.h (#95)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "a78acb2d2ca07f51ce33b3995e9b45bfb691e455", "messageHeadline": "rename graph.h to graph_viewer.h (#84)", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 42, "deletions": 42}}, {"node": {"oid": "980e8aa270bcd9ec79b0c429bfe7c3f21e9d4f48", "messageHeadline": "Update BUILD.md (#90)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 40, "deletions": 2}}, {"node": {"oid": "744ec28c88eee3692806105a64c518fb6169698d", "messageHeadline": "(de)quantize_linear optimizations (#76)", "author": {"name": "Brian Martin", "email": "42186431+martinb35@users.noreply.github.com", "user": {"login": "martinb35"}}, "additions": 195, "deletions": 97}}, {"node": {"oid": "37385ec02980f62f3861fcb9b3b54576122d41c9", "messageHeadline": "Add Dockerfile and page for ARM builds. (#83)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 41, "deletions": 2}}, {"node": {"oid": "1e59b6f1c2c8ec6bb2d3012341764aa86a93db41", "messageHeadline": "Minor fixes to CI definitions (#80)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 3, "deletions": 2}}, {"node": {"oid": "47a6992e1b2b8974aceeecbba46e2e061ea6fc0d", "messageHeadline": "update mkldnn to 0.17.1 and address assumptions related to tensor pad\u2026", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 78, "deletions": 162}}, {"node": {"oid": "900e69ceae1e6d3ebf6dffd47729eb3c43baf647", "messageHeadline": "User lower case while comparing the activation method as it's not cle\u2026", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 74, "deletions": 125}}, {"node": {"oid": "f1c66a4aae123790c894b64ed527078f27156814", "messageHeadline": "Updated CSharp package description (#75)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1c9d0b27293019eda0bbc189da2ad6fc5a0aef2e", "messageHeadline": "Add missing types for Slice op (#74)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 77, "deletions": 11}}, {"node": {"oid": "aa549cd194aa719bdb7231d1c766296b37e5bd91", "messageHeadline": "Support fusing 3D Conv with Add/Mul. (#23)", "author": {"name": "Weixing Zhang", "email": "weixingzhang@users.noreply.github.com", "user": {"login": "weixingzhang"}}, "additions": 208, "deletions": 41}}, {"node": {"oid": "b7cc611563cfd1bafdff14e38fb50ec9c48c3d68", "messageHeadline": "Minor documentation changes (#78)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 15, "deletions": 14}}, {"node": {"oid": "07253fd64732134e2c64c1917fccea9cf60c6c69", "messageHeadline": "Faxu patch 1 (#63)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 30, "deletions": 13}}, {"node": {"oid": "21dad1780b597253d318b456f9040bd0bf9c80ea", "messageHeadline": "Add logo and link to C# pkg. (#73)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 3, "deletions": 1}}, {"node": {"oid": "28bb96be2a8daf2d27ba4c3b263f26350a43a340", "messageHeadline": "use std::copy to avoid gsl::span overhead (#71)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "fd0d7c5fc050d8d87bb207289d4db219a9618ccb", "messageHeadline": "Add the mac python packaging script (#72)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 47, "deletions": 0}}, {"node": {"oid": "31780cae32c3d234a993a3f343939eeac35931f3", "messageHeadline": "Add NonMaxSupression op to contribution ops (#60)", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 499, "deletions": 0}}, {"node": {"oid": "6624dd277815cf57c70e75e28aadf0a2f5cd97bc", "messageHeadline": "Rel 0.1.5 (#70)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 11, "deletions": 5}}, {"node": {"oid": "21044846eefc0720e0c52bc075e4e701fcc2acc6", "messageHeadline": "use #pragma parallel for properly (#68)", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 2, "deletions": 3}}, {"node": {"oid": "c55663860e96799c1c9af9d155c31b178c52a633", "messageHeadline": "fixed metadata element -- use PackageProjectUrl instead of ProjectUrl\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "b534f9fa5f9e72275a8a5d4fef8710bdc43d9a05", "messageHeadline": "Square and Cube optimization for Pow<float>() operator. (#30)", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 14, "deletions": 1}}, {"node": {"oid": "bcc8f621ea7b2ae6785faa979e9f34c2b4c39ccd", "messageHeadline": "updated nuget package metadata for MS compliance (#66)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "edbe19d22fc967e9c31f128dee0d228e0496829c", "messageHeadline": "Include output name with mismatch error message so it's clearer where\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "9b1bc06c348e298c235a1c421e8ea08f4709047c", "messageHeadline": "Merge pull request #64 from Microsoft/kezhan/remove_const_cast", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 55, "deletions": 47}}, {"node": {"oid": "e9de9abc961e2ea09ac0550239cd8c8f083e6239", "messageHeadline": "update", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 7, "deletions": 8}}, {"node": {"oid": "fbd7b9b8f0dca24789c8b0e54e8f1f39eebb2afb", "messageHeadline": "Merge branch 'master' of https://github.com/Microsoft/onnxruntime int\u2026", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "56811d10aff166f8c002b49fd99d353b19115e0b", "messageHeadline": "remove const cast in transformer_memcpy", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 37, "deletions": 32}}, {"node": {"oid": "e00c95625480f6c29e4a2b973ca2a2256f834a78", "messageHeadline": "update the comments.", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "39ebccbc8b2c0285dc5fe5d064bf99bf68b52aa1", "messageHeadline": "Fix sample example documentation for python pkg (#61)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "afe3fb5c1dd076e4a6d5736bd79549e4439cbd40", "messageHeadline": "Merge branch 'master' of https://github.com/Microsoft/onnxruntime int\u2026", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 21, "deletions": 5}}, {"node": {"oid": "35f94cac279e63edda35b726bc15e0f45c7e9298", "messageHeadline": "remove const cast from conv related fusion.", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 11, "deletions": 7}}, {"node": {"oid": "1ea32a097a362ab98d7e3a7914042c8d03293336", "messageHeadline": "Update pythin bindings for options (#14)", "author": {"name": "Xavier Dupr\u00e9", "email": "xadupre@users.noreply.github.com", "user": {"login": "xadupre"}}, "additions": 21, "deletions": 5}}, {"node": {"oid": "728c3c078ecb564deedaee003f240769304be978", "messageHeadline": "Update issue templates (#62)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 49, "deletions": 0}}, {"node": {"oid": "cd1042c94fcd4deceac3c02f39cd10a1f16ee62e", "messageHeadline": "Merge pull request #36 from Microsoft/zhalei/softmax_optimize", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 50, "deletions": 8}}, {"node": {"oid": "fafc48bf94248900cd90222a731bedd04666bcf2", "messageHeadline": "Merge pull request #59 from Microsoft/kezhan/qlinearmatmul", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 27, "deletions": 2}}, {"node": {"oid": "91c860bec408152ef039fa5e5dd7a5f82ed52f64", "messageHeadline": "fix description", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 9, "deletions": 9}}, {"node": {"oid": "34aa545b64940fd9f48213d0a8b0ba341c77891f", "messageHeadline": "add QLinearMatMul", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 27, "deletions": 2}}, {"node": {"oid": "e8db06ed44bed2f2aacdc21bd332d5955ef93527", "messageHeadline": "Merge pull request #32 from Microsoft/kezhan/op_spec_fix", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "7780fd6cd9e4f9b6a20f65488a7b24bcfb75bc8e", "messageHeadline": "fix bug -- failing on non float tensor output types (#58)", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 300, "deletions": 4}}, {"node": {"oid": "8980cbbfada0f14ac53330e983acfe61d21d821a", "messageHeadline": "Merge pull request #52 from Microsoft/scmckay/UpdateCudaInfoInBuildMd", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 47, "deletions": 27}}, {"node": {"oid": "a4bcb1121b76901186dbf7fe23d1571f942dc225", "messageHeadline": "Merge pull request #57 from Microsoft/scmckay/FixInferenceSessionInpu\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 170, "deletions": 26}}, {"node": {"oid": "97dc949e3f0bdd0eb79da68750f02f3699ab8feb", "messageHeadline": "Fix some more C# unit test issues.", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 5, "deletions": 3}}, {"node": {"oid": "f3fb1d5ff8ce6c104d76981ced59ea85f8c2fa86", "messageHeadline": "Merge remote-tracking branch 'origin/master' into scmckay/FixInferenc\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "2486336355c5a0af23ecd0b63a16251f514eb675", "messageHeadline": "Update error messages C# tests check for.", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "d263d6734e97029bd79e211de6150a757023db92", "messageHeadline": "Adjust test. valid input names aren't deterministically ordered in er\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 0, "deletions": 1}}, {"node": {"oid": "4dd1e50aa7e59683c05f107bfd019dd238c8b3cd", "messageHeadline": "Merge remote-tracking branch 'origin/master' into scmckay/UpdateCudaI\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 94, "deletions": 41}}, {"node": {"oid": "ca86d8f17c32b7a9d86c016e14c5e2b2f8162e3b", "messageHeadline": "Merge pull request #55 from Microsoft/jywu/incr_fixup", "author": {"name": "jywu-msft", "email": "43355415+jywu-msft@users.noreply.github.com", "user": {"login": "jywu-msft"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "ea1ad0aaf541d42a8384bdbd82b9addc41a8635d", "messageHeadline": "Merge remote-tracking branch 'origin/master' into scmckay/FixInferenc\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 90, "deletions": 38}}, {"node": {"oid": "989b00321e9a2bc2043764622dc84b6ed1aeee03", "messageHeadline": "Update session state initializer to support overriding initializers.", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 10, "deletions": 8}}, {"node": {"oid": "3feeb9de3e3121879d797bd2e838a0adf7b211f4", "messageHeadline": "Merge branch 'master' of https://github.com/Microsoft/onnxruntime int\u2026", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 113, "deletions": 58}}, {"node": {"oid": "846044e28289defe2408aca22bebd1710f582f30", "messageHeadline": "[Mac] support mkldnn for macOS (#56)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 12, "deletions": 3}}, {"node": {"oid": "7523e76649b8d6ba4610f7f6532bd73fd04f6b2b", "messageHeadline": "Minor wording changes to design doc (#51)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 17, "deletions": 18}}, {"node": {"oid": "6371025860dada8dc30671ac148cdb5b179a5698", "messageHeadline": "Add flag for mac compliance (#45)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 7, "deletions": 0}}, {"node": {"oid": "d60507d2e9ea5f55afb72e864195b62a580da464", "messageHeadline": "[Mac] fix python binding (#54)", "author": {"name": "Yulong Wang", "email": "f.s@qq.com", "user": {"login": "fs-eire"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "91f1866b648b144a0d7c49e26f047c7b95493c1b", "messageHeadline": "Only check for 'bin' folder in cudnn path on windows.", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 6, "deletions": 5}}, {"node": {"oid": "bfaade660be6da050150741bc7d24456524e3841", "messageHeadline": "Initial changes.", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 157, "deletions": 16}}, {"node": {"oid": "6179ddb2a14eefb159ceab9bc8f058ac9345f2c8", "messageHeadline": "fix for possible incremental build breaks.", "author": {"name": "George Wu", "email": "jywu@microsoft.com", "user": {"login": "jywu-msft"}}, "additions": 4, "deletions": 3}}, {"node": {"oid": "cb1781927fae1d2840f4554324d4d6a73e14370d", "messageHeadline": "Remove mklml in Linux python wheel packagaing (#53)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 2, "deletions": 2}}, {"node": {"oid": "21cabe3ead4bd238a1e7b3b1c6ff7659e5b177a8", "messageHeadline": "Merge pull request #35 from Microsoft/zhalei/reduce_mean_optimize", "author": {"name": "Zhang Lei", "email": "zhang.huanning@hotmail.com", "user": {"login": "zhanghuanrong"}}, "additions": 49, "deletions": 12}}, {"node": {"oid": "2f619263900e488a51f8d833f9b01762d2a9c67e", "messageHeadline": "Update build.md section on CUDA builds to clarify and add CUDA 10.0 i\u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 46, "deletions": 27}}, {"node": {"oid": "2f234e4e7864dfddf07d9012ebdf357ed1efeab9", "messageHeadline": "Minor fixes to nuget README.md file (#49)", "author": {"name": "jignparm", "email": "jignparm@microsoft.com", "user": {"login": "jignparm"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "564907fa1a5b20c5742c3322e63f40877f7514f2", "messageHeadline": "remove the non ascii char (#48)", "author": {"name": "Tang, Cheng", "email": "souptc@gmail.com", "user": {"login": "souptc"}}, "additions": 7, "deletions": 7}}, {"node": {"oid": "16ef49c51b6f16f22c5b92b958a63c85b084fd6a", "messageHeadline": "Merge pull request #47 from Microsoft/dev/shahasad/some-cosmetic-fixe\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 10, "deletions": 7}}, {"node": {"oid": "0540d8e5f71835e836a188cf6ce0de7dd23b2d81", "messageHeadline": "Better name for read.", "author": {"name": "Lei Zhang", "email": "zhalei@microsoft.com", "user": null}, "additions": 8, "deletions": 8}}, {"node": {"oid": "7cb3dfc18a043fe1d9c402cfa75024150742e431", "messageHeadline": "Optimize ReduceMean/ReduceSum when all reduce axises located at the t\u2026", "author": {"name": "Lei Zhang", "email": "zhalei@microsoft.com", "user": null}, "additions": 49, "deletions": 12}}, {"node": {"oid": "6f65a03939251e3f966410a225affe9299e1f567", "messageHeadline": "some cleanup", "author": {"name": "Shah Asaduzzaman", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 10}}, {"node": {"oid": "c1613aa28e35f3ecf5fccfa88591573e1b6a1a63", "messageHeadline": "Fixes in the NuGet package meatada. Include sourcelink", "author": {"name": "Shah Asaduzzaman", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 17, "deletions": 6}}, {"node": {"oid": "17b72b2e7dca1579cd75a3dc17ff352fb9f905d3", "messageHeadline": "Merge pull request #46 from Microsoft/tracysh/mergemaster", "author": {"name": "Tracy Sharpe", "email": "42477615+tracysh@users.noreply.github.com", "user": {"login": "tracysh"}}, "additions": 163, "deletions": 82}}, {"node": {"oid": "39fc17281ce8f5520f67d0d1f18a747cd04858aa", "messageHeadline": "merge from internal master", "author": {"name": "Tracy Sharpe", "email": "tracysh@microsoft.com", "user": {"login": "tracysh"}}, "additions": 163, "deletions": 82}}, {"node": {"oid": "6b00e6bb4df23e77da73b72c67ff98ba605712f1", "messageHeadline": "Simpler unused parameter in #if defined() switch.", "author": {"name": "Lei Zhang", "email": "zhalei@microsoft.com", "user": null}, "additions": 3, "deletions": 7}}, {"node": {"oid": "c530064ebe912621ac5627bc486c62356fbd1eeb", "messageHeadline": "Better opemmp parallel group count calculation in Softmax parallel ru\u2026", "author": {"name": "Lei Zhang", "email": "zhalei@microsoft.com", "user": null}, "additions": 37, "deletions": 18}}, {"node": {"oid": "bd50598d17e268b515b94ce9e45215fe8260e9d5", "messageHeadline": "Document the Graph header files and cleanup some issues. (#42)", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 965, "deletions": 682}}, {"node": {"oid": "4fbd3457327e589229ac21efae52256432a24fe4", "messageHeadline": "Merge pull request #44 from Microsoft/fix_codeowners_file", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "75d329d84525ba31e588b09e40c4c16bf43dd4f4", "messageHeadline": "Merge pull request #25 from Microsoft/dev/shahasad/merge-latest-few-c\u2026", "author": {"name": "shahasad", "email": "43590019+shahasad@users.noreply.github.com", "user": {"login": "shahasad"}}, "additions": 661, "deletions": 1525}}, {"node": {"oid": "6e2a1ceb41f0f50b1bc49281baa34b13d8a67c08", "messageHeadline": "merged from master", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 241, "deletions": 57}}, {"node": {"oid": "be1be9fa942b4592df182e7c1595fc52cbc48edc", "messageHeadline": "explicitly specify outputpath in the test project", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "f638031f02a108a4e1f19587b18ab8be21b8c7d7", "messageHeadline": "make test explicitly anyCPU", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1a03f449381bbcda014e11869aee2c7d6edf3f15", "messageHeadline": "make the test project explicitly x64 platform to get working", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "6038ba21fd670dbfcd55354b876ac1d4f8161b09", "messageHeadline": "Fix codeowners file", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "1b3efc36c10a7ca5e4ed20f8c3f3c1156c764219", "messageHeadline": "Add pipeline for building python wheels (#41)", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 189, "deletions": 1}}, {"node": {"oid": "3f4589ced53a876a2618d9583721ba5d7fa3a0fc", "messageHeadline": "Add remaining build options and make minor changes in documentation (\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 6, "deletions": 6}}, {"node": {"oid": "29f9e89226a2e6bb7ba0287618f4dcf6ffe9f5a3", "messageHeadline": "merged the latest csharp folder changes from VSTS repo @313681f9", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 268, "deletions": 9}}, {"node": {"oid": "d7d43bc13fb52843231d15f5cdc087611fea929a", "messageHeadline": "merged from master", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 736, "deletions": 140}}, {"node": {"oid": "3b1140954e8c621c1bce70685c411050a51bc235", "messageHeadline": "Bug bash (#43)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 46, "deletions": 50}}, {"node": {"oid": "a9b52f399d9f04f72e330c35cc3fa2e84027317f", "messageHeadline": "Add pre-release notice to c_api (#38)", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 4, "deletions": 0}}, {"node": {"oid": "c07b4aff5e3affd3259c9684e303c6fc84242be5", "messageHeadline": "Merge pull request #37 from Microsoft/ryanunderhill/hide_leak", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 12, "deletions": 9}}, {"node": {"oid": "047ca7e57e097c742a9396a155ec1df9f6297653", "messageHeadline": "Merge branch 'master' into ryanunderhill/hide_leak", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 446, "deletions": 106}}, {"node": {"oid": "df1d01f85349540177725ea2615b06a499bc1bdc", "messageHeadline": "Update CI configs to test mkldnn", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 1, "deletions": 1}}, {"node": {"oid": "dcd4e0cb8fe839d6693a2d92f694d4feecba3a41", "messageHeadline": "Change debug_alloc to not exit(-1) on the command line so that our bu\u2026", "author": {"name": "RyanUnderhill", "email": "ryanhill@microsoft.com", "user": {"login": "RyanUnderhill"}}, "additions": 11, "deletions": 8}}, {"node": {"oid": "9c4020678c91048f03281036e4cba768713c6187", "messageHeadline": "Merge pull request #21 from Microsoft/duli/shape_inference", "author": {"name": "Du Li", "email": "du.li@microsoft.com", "user": {"login": "duli2012"}}, "additions": 32, "deletions": 4}}, {"node": {"oid": "725a262cebf564aa72a474569ea85c3d728bf7eb", "messageHeadline": "Merge branch 'duli/shape_inference' of github.com:Microsoft/onnxrunti\u2026", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "e75bde2e975647c0ccb617900125cc64a0943aed", "messageHeadline": "Accomodating PR comments.", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 0, "deletions": 2}}, {"node": {"oid": "e7bdfa00db1d0e9909f1d0d5e159173e3fbd608f", "messageHeadline": "Optimize softmax cpu by parallel using openmp.", "author": {"name": "Lei Zhang", "email": "zhalei@microsoft.com", "user": null}, "additions": 34, "deletions": 7}}, {"node": {"oid": "408fd21a7ff293e4ed4d62675576b8f2e63b60c0", "messageHeadline": "Windows CI: enable pybind (#34)", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4fe704521462252754fbe1fe5df3f74db9711db9", "messageHeadline": "Faxu documentation (#16)", "author": {"name": "Faith Xu", "email": "txsafx@gmail.com", "user": {"login": "faxu"}}, "additions": 54, "deletions": 20}}, {"node": {"oid": "ab4d4e1ae09459fac42536cfc2766ff02fee1453", "messageHeadline": "fix  output type specification for ConvInteger", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "0bb5cb7a3075905f6d9894151b76f4dffdf295c4", "messageHeadline": "Accomodating PR comments.", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "573a9aba3c1ccdcdfa6ee93a4c93e2ac49323efd", "messageHeadline": "Merge pull request #29 from Microsoft/RyanUnderhill-patch-1", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 13, "deletions": 15}}, {"node": {"oid": "6982325b292e4a5ee84ed43974724b7dd8f00ce9", "messageHeadline": "Merge pull request #22 from Microsoft/scmckay/SupportUnknownDimension\u2026", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 346, "deletions": 66}}, {"node": {"oid": "3ccb0876fd86808c8bd0200d00ac995d959047a7", "messageHeadline": "Merge", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 275, "deletions": 26}}, {"node": {"oid": "3761d2d718b9147670d3a5c454b5e998d6b1baa2", "messageHeadline": "Update C_API.md", "author": {"name": "Ryan Hill", "email": "38674843+RyanUnderhill@users.noreply.github.com", "user": {"login": "RyanUnderhill"}}, "additions": 13, "deletions": 15}}, {"node": {"oid": "194cc1539c0640516db3f968d56339e7bbacb3ee", "messageHeadline": "Merge pull request #24 from Microsoft/scmckay/Scan_SupportScalarInputs", "author": {"name": "Scott McKay", "email": "Scott.McKay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 51, "deletions": 26}}, {"node": {"oid": "930145f2d217163b01fca08d8236125137897370", "messageHeadline": "add mkldnn and openmp on windows CI build. C# package binary includes\u2026", "author": {"name": "Shah Asaduzzaman", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "84fa1018a3fb939941775e8f6f88f41fb114beb4", "messageHeadline": "Create CODEOWNERS (#27)", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 0}}, {"node": {"oid": "b9dc15310694e917a170951a98d0296ba5272ceb", "messageHeadline": "Merge pull request #20 from linkerzhang/kezhan/quantize_ops", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 223, "deletions": 0}}, {"node": {"oid": "e477e7563b722ff54db75f24e6688c2f5bff7163", "messageHeadline": "typo in last commit", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "4f0ec77f6e77017336a6a929403ba1f59769acec", "messageHeadline": "Merge branch 'dev/shahasad/enable-dll-and-csharp-in-windows-ci-build'\u2026", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "33537055cbd092d459bb2010cc70fba5bf53cc32", "messageHeadline": "make mkldnn.dll optional", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "d8cf3f0e33d18b06b1cb7742b5fa2ccd44af325f", "messageHeadline": "enabled shared-lib and csharp build in windows CI", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "ba6123db57f2e868fdd037b2a73228a451cdd5c8", "messageHeadline": "fix the exceptions", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 7, "deletions": 16}}, {"node": {"oid": "41e1ed8e0196b970e96406be853d6b100cb8f3c3", "messageHeadline": "added the CSHarp doc", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 115, "deletions": 1}}, {"node": {"oid": "f454826b932a5131321426a4db70bfcd1a30440c", "messageHeadline": "updated the files with the latest ones from Lotus VSTS repo", "author": {"name": "Shah Asaduzzaman (ASAD)", "email": "shahasad@microsoft.com", "user": {"login": "shahasad"}}, "additions": 277, "deletions": 1507}}, {"node": {"oid": "eef4db37f335c5f2af220b4019f29c7eb33a59e5", "messageHeadline": "Fix a couple of warnings from the linux and VC 14.11 toolset builds", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 12, "deletions": 2}}, {"node": {"oid": "03d7d25989dc21d08ec2aa70632483092bfdce98", "messageHeadline": "Support scalars (zero dimensions) in Scan by allowing the parameters \u2026", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 51, "deletions": 26}}, {"node": {"oid": "720aca581a2f2b918b61d7f609848100b35c58f6", "messageHeadline": "Update comments", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 16, "deletions": 12}}, {"node": {"oid": "eef36d2fbf14986472348e6be2b4d9e5132d3f28", "messageHeadline": "Update some comments", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 9, "deletions": 2}}, {"node": {"oid": "e7e801b45e1a2844f5230f4b7fd9fa3340036c63", "messageHeadline": "Adding shaper inference for Op expand_dims", "author": {"name": "Du Li", "email": "duli1@microsoft.com", "user": {"login": "duli2012"}}, "additions": 34, "deletions": 4}}, {"node": {"oid": "5d3992f999ec8bc2d30ddf9ba8686b3054bce595", "messageHeadline": "Handle the Scan subgraph producing outputs with a symbolic dimension.", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 325, "deletions": 66}}, {"node": {"oid": "c7513e676fa7ac0b84a2462891344bc9d82eef8b", "messageHeadline": "Merge pull request #19 from Microsoft/fix_build_status_badge", "author": {"name": "Ke Zhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "a82a0907e0afda1b0bd623792699e2a7cd0dc7eb", "messageHeadline": "add ops for quantization support.", "author": {"name": "linkerzhang", "email": "kezhan@microsoft.com", "user": {"login": "linkerzhang"}}, "additions": 223, "deletions": 0}}, {"node": {"oid": "d35bea7455d4064eeccf97636500969d74fcc2eb", "messageHeadline": "Fix build status badge.", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "faa06d3516f2b64758697908bb5d1f2d703911f3", "messageHeadline": "Merge pull request #17 from Microsoft/scmckay/AllowForOptionalInputsI\u2026", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "acf9aea4aa4ba43cb69ade849809de38c9215e00", "messageHeadline": "Merge pull request #13 from Microsoft/chasun/p1", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 0, "deletions": 0}}, {"node": {"oid": "b3c5da5b97841906fde4ccd72614ce0ac67369e9", "messageHeadline": "Merge pull request #18 from Microsoft/sync_internal_master", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 2460, "deletions": 1345}}, {"node": {"oid": "7aef8a1ccab9c0ba7f6f47f6c5237fc071d870b8", "messageHeadline": "Sync with internal master.", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 2460, "deletions": 1345}}, {"node": {"oid": "0ee3ad32d2043f171aead47ad1384d039a4ecba6", "messageHeadline": "Allow for optional graph inputs in InferenceSession.run", "author": {"name": "Scott McKay", "email": "scmckay@microsoft.com", "user": {"login": "skottmckay"}}, "additions": 2, "deletions": 1}}, {"node": {"oid": "ba470c21808783824791af80eec7ed391b5c867c", "messageHeadline": "Merge pull request #12 from Microsoft/update_onnx_commit", "author": {"name": "Changming Sun", "email": "chasun@microsoft.com", "user": {"login": "snnn"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "f4ea04870ec2a73b613e6b39c0d3645f7fbb3c20", "messageHeadline": "delete extra code in onnxruntime_unittests.cmake", "author": {"name": "Changming Sun", "email": "me@sunchangming.com", "user": {"login": "snnn"}}, "additions": 0, "deletions": 4}}, {"node": {"oid": "9fbcb7fa51d3d8984422c326dbed54e5eb91c732", "messageHeadline": "Update onnx to commit to pickup shape inference fix for broadcast, co\u2026", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 1, "deletions": 1}}, {"node": {"oid": "35b43cfebc48106a37ca1c8a56f6a975168dedcc", "messageHeadline": "Merge pull request #10 from Microsoft/hecli/cuda9.1", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 10, "deletions": 13}}, {"node": {"oid": "d2a8b0e65b8bdaa17fd496a25b2e2a5f114ac5f8", "messageHeadline": "update the cuda cudnn version information", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 2, "deletions": 2}}, {"node": {"oid": "c2d5b7e368b7d2cd1b16b885c71444eb0914b30c", "messageHeadline": "update linux gpu to use cuda 9.1 + cudnn 7.1.2", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 6, "deletions": 9}}, {"node": {"oid": "8f716cb4fc78f783b51dcc10f05b9c7b85e94129", "messageHeadline": "Switch GPU build to use CUDA 9.1 + cudnn 7.0.5", "author": {"name": "Hector Li", "email": "hecli@microsoft.com", "user": {"login": "HectorSVC"}}, "additions": 3, "deletions": 3}}, {"node": {"oid": "0f96064e7029403331a18f778af4bb8eb288729b", "messageHeadline": "Merge pull request #9 from Microsoft/mac-ci", "author": {"name": "Pranav Sharma", "email": "emailpranav@gmail.com", "user": {"login": "pranavsharma"}}, "additions": 15, "deletions": 1}}, {"node": {"oid": "53ce3e7b55ceb2b431fceb3539482acda4efc72b", "messageHeadline": "Enable Mac pipeline", "author": {"name": "Raymond Yang", "email": "zihao.yang@microsoft.com", "user": null}, "additions": 15, "deletions": 1}}, {"node": {"oid": "89618e8f1e8fee0354de57494bb71da96a445b71", "messageHeadline": "Initial bootstrap commit.", "author": {"name": "Pranav Sharma", "email": "prs@microsoft.com", "user": {"login": "pranavsharma"}}, "additions": 127587, "deletions": 0}}]
\ No newline at end of file