12350 commits

This branch

This branch

All branches

Author	SHA1	Message	Date
Jian Chen	cf2ba46ce4	Change hostArchitecture to arm64 for aarch64	2025-02-07 15:10:33 -08:00
Jian Chen	1176c96540	Change hostArchitecture to arm64 for aarch64	2025-02-07 15:08:03 -08:00
Jian Chen	3737846fae	Change the build artifact to pipeline artifact.	2025-02-07 08:49:42 -08:00
Jian Chen	bb75778c7a	Update os	2025-02-06 21:55:58 -08:00
Jian Chen	16c8269520	Update os	2025-02-06 19:36:09 -08:00
Jian Chen	15347b8841	Update os	2025-02-06 19:31:06 -08:00
Jian Chen	67c2221a32	Adding a new stage	2025-02-06 19:25:40 -08:00
Jian Chen	071463027f	Adding a new stage	2025-02-06 19:22:56 -08:00
Jian Chen	58183eb1a1	Adding a new stage	2025-02-06 19:17:49 -08:00
Jian Chen	293d96f915	undo templates	2025-02-06 19:11:40 -08:00
Jian Chen	8705d205f2	undo pool	2025-02-06 19:10:09 -08:00
Jian Chen	7d2e988d02	undo binary	2025-02-06 19:08:48 -08:00
Jian Chen	c17734e624	onnxruntime-inference-examples	2025-02-06 19:06:16 -08:00
Jian Chen	5108e87b33	undo dml-vs-2022.yml	2025-02-06 19:02:41 -08:00
Jian Chen	979a1a78a1	undo dml-vs-2022.yml	2025-02-06 19:01:25 -08:00
Jian Chen	3d8da4d0b0	undo Binary c-api-cpu.yml 1	2025-02-06 18:59:24 -08:00
Jian Chen	645f2370ab	undo Binary c-api-cpu.yml 1.25 ... 1ES.PublishPipelineArtifact@0 to 1ES.PublishPipelineArtifact@1	2025-02-06 18:58:00 -08:00
Jian Chen	abb684873e	undo Binary c-api-cpu.yml 1.25	2025-02-06 18:56:10 -08:00
Jian Chen	73b7200ffe	undo Binary c-api-cpu.yml 1.5	2025-02-06 18:55:00 -08:00
Jian Chen	bcb0652298	undo Binary c-api-cpu.yml 1.75	2025-02-06 18:53:08 -08:00
Jian Chen	9746f1cb9b	undo Binary c-api-cpu.yml 1.5	2025-02-06 18:51:58 -08:00
Jian Chen	791304291f	undo Binary c-api-cpu.yml 1	2025-02-06 18:50:18 -08:00
Jian Chen	503b159f18	undo Binary c-api-cpu.yml 2	2025-02-06 18:49:15 -08:00
Jian Chen	073a6c5675	undo Binary c-api-cpu.yml 3	2025-02-06 18:48:07 -08:00
Jian Chen	b16aac39a2	undo Binary c-api-cpu.yml 3	2025-02-06 18:47:35 -08:00
Jian Chen	4bdec8e0bf	Binary c-api-cpu.yml 3	2025-02-06 18:45:32 -08:00
Jian Chen	27eadb158b	Binary c-api-cpu.yml 2	2025-02-06 18:44:52 -08:00
Jian Chen	10355d77e3	Binary c-api-cpu.yml 2	2025-02-06 18:44:14 -08:00
Jian Chen	165d968fb7	Binary c-api-cpu.yml 2	2025-02-06 18:43:55 -08:00
Jian Chen	faa38bffbf	Binary c-api-cpu.yml	2025-02-06 18:41:35 -08:00
Jian Chen	de9ce655fc	Disable c-api-cpu.yml	2025-02-06 18:40:03 -08:00
Jian Chen	1b3dcc89fe	Update c-api-cpu.yml	2025-02-06 18:39:15 -08:00
Jian Chen	41f2aa32e1	Disable java-cuda-packaging-stage.yml	2025-02-06 18:34:04 -08:00
Jian Chen	471f287235	Disable c-api-cpu.yml	2025-02-06 18:33:11 -08:00
Jian Chen	77110697f4	Disable nuget-combine-cuda-stage.yml	2025-02-06 18:32:11 -08:00
Jian Chen	05dae73477	Disable dml	2025-02-06 18:30:52 -08:00
Jian Chen	7e22fb64cb	1ES	2025-02-06 18:27:42 -08:00
Jian Chen	39ab30674d	publish	2025-02-06 18:24:30 -08:00
Jian Chen	6e45a7bf1d	Try to skip validate-package.yml	2025-02-06 18:22:40 -08:00
Jian Chen	f9aa616b04	Try to skip validate-package.yml	2025-02-06 18:20:53 -08:00
Jian Chen	60749a2e5f	Try to skip validate-package.yml	2025-02-06 18:17:58 -08:00
Jian Chen	5cf3f47137	Try to skip validate-package.yml	2025-02-06 18:16:42 -08:00
Jian Chen	361c41ed0a	Try to skip ESRP	2025-02-06 18:15:27 -08:00
Jian Chen	9efa0b4965	Migrate Zip-Nuget Package Pipeline to 1ES	2025-02-06 18:07:35 -08:00
Jian Chen	33e6ebfe2f	Migrate Zip-Nuget Package Pipeline to 1ES	2025-02-06 17:59:23 -08:00
Jian Chen	0469e1577d	Migrate Zip-Nuget Package Pipeline to 1ES	2025-02-06 17:58:54 -08:00
Jian Chen	d9dda06456	Migrate Zip-Nuget Package Pipeline to 1ES	2025-02-06 17:57:59 -08:00
Jian Chen	702ed1ce0f	Migrate Zip-Nuget Package Pipeline to 1ES	2025-02-06 17:56:34 -08:00
microsoft-github-policy-service[bot]	65008cbb73	Auto-generated baselines by 1ES Pipeline Templates (#23603)	2025-02-06 17:06:29 -08:00
Tianlei Wu	09e5724f3b	[CUDA] Fix beam search of num_beams > 32 (#23599) ... ### Description * Pass topk_scores to beam scorer in slow topk path. * Add an env variable `ORT_BEAM_SEARCH_USE_FAST_TOPK` to enable/disable fast topk. * Add a test case for slow topk path. ### Motivation and Context This bug was introduced in https://github.com/microsoft/onnxruntime/pull/16272 Beam search uses fast cuda kernel when number of beams <= 32. When beam size is larger than that threshold, we use another code path (slower cuda kernel) to get topk. In such `slow topk path`, topk_scores shall be passed to beam scorer but it is not. This bug will cause incorrect result when num_beams > 32. It was not found previously since such large beam size is rarely used.	2025-02-06 16:50:31 -08:00