TensorRT-LLM icon indicating copy to clipboard operation
TensorRT-LLM copied to clipboard

[Feat] add chunked-attention kernels on Hopper (for llama4)

Open PerkzZheng opened this issue 6 months ago • 28 comments

Add chunked-attention kernels on Hopper (for llama4)

fmha_v2 commit: 6552b99d4820fa3f5e8a48a392681a8c128bf623

Description

Please explain the issue and the solution in short.

Test Coverage

GitHub Bot Help

/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...

Provide a user friendly way for developers to interact with a Jenkins server.

Run /bot [-h|--help] to print this help message.

See details below for each supported subcommand.

run [--disable-fail-fast --skip-test --stage-list "A10-1, xxx" --gpu-type "A30, H100_PCIe" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-[Post-Merge]-1, xxx"]

Launch build/test pipelines. All previously running jobs will be killed.

--disable-fail-fast (OPTIONAL) : Disable fail fast on build/tests/infra failures.

--skip-test (OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.

--stage-list "A10-1, xxx" (OPTIONAL) : Only run the specified test stages. Examples: "A10-1, xxx". Note: Does NOT update GitHub check status.

--gpu-type "A30, H100_PCIe" (OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.

--only-multi-gpu-test (OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.

--disable-multi-gpu-test (OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.

--add-multi-gpu-test (OPTIONAL) : Force run the multi-GPU tests. Will also run L0 pre-merge pipeline.

--post-merge (OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.

--extra-stage "H100_PCIe-[Post-Merge]-1, xxx" (OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-[Post-Merge]-1, xxx".

kill

kill

Kill all running builds associated with pull request.

skip

skip --comment COMMENT

Skip testing for latest commit on pull request. --comment "Reason for skipping build/test" is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

reuse-pipeline

reuse-pipeline

Reuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.

PerkzZheng avatar May 14 '25 07:05 PerkzZheng

/bot run --disable-fail-fast

PerkzZheng avatar May 14 '25 07:05 PerkzZheng

/bot run --disable-fail-fast

PerkzZheng avatar May 14 '25 07:05 PerkzZheng

PR_Github #5135 [ run ] triggered by Bot

tensorrt-cicd avatar May 14 '25 07:05 tensorrt-cicd

PR_Github #5135 [ run ] completed with state FAILURE /LLM/main/L0_MergeRequest_PR pipeline #3745 completed with status: 'FAILURE'

tensorrt-cicd avatar May 14 '25 09:05 tensorrt-cicd

/bot run --disable-fail-fast

PerkzZheng avatar May 14 '25 11:05 PerkzZheng

PR_Github #5168 [ run ] triggered by Bot

tensorrt-cicd avatar May 14 '25 11:05 tensorrt-cicd

/bot run --disable-fail-fast

PerkzZheng avatar May 14 '25 14:05 PerkzZheng

PR_Github #5186 [ run ] triggered by Bot

tensorrt-cicd avatar May 14 '25 14:05 tensorrt-cicd

PR_Github #5168 [ run ] completed with state ABORTED

tensorrt-cicd avatar May 14 '25 14:05 tensorrt-cicd

PR_Github #5186 [ run ] completed with state SUCCESS /LLM/main/L0_MergeRequest_PR pipeline #3784 completed with status: 'FAILURE'

tensorrt-cicd avatar May 14 '25 22:05 tensorrt-cicd

/bot run --disable-fail-fast

PerkzZheng avatar May 15 '25 00:05 PerkzZheng

PR_Github #5223 [ run ] triggered by Bot

tensorrt-cicd avatar May 15 '25 00:05 tensorrt-cicd

PR_Github #5223 [ run ] completed with state SUCCESS /LLM/main/L0_MergeRequest_PR pipeline #3815 completed with status: 'SUCCESS'

tensorrt-cicd avatar May 15 '25 06:05 tensorrt-cicd

/bot run --disable-fail-fast

PerkzZheng avatar May 15 '25 10:05 PerkzZheng

PR_Github #5337 [ run ] triggered by Bot

tensorrt-cicd avatar May 15 '25 10:05 tensorrt-cicd

/bot run --disable-fail-fast

PerkzZheng avatar May 15 '25 13:05 PerkzZheng

PR_Github #5354 [ run ] triggered by Bot

tensorrt-cicd avatar May 15 '25 13:05 tensorrt-cicd

PR_Github #5337 [ run ] completed with state ABORTED

tensorrt-cicd avatar May 15 '25 13:05 tensorrt-cicd

/bot run --disable-fail-fast

mikeiovine avatar May 15 '25 17:05 mikeiovine

PR_Github #5354 [ run ] completed with state FAILURE /LLM/main/L0_MergeRequest_PR pipeline #3907 completed with status: 'FAILURE'

tensorrt-cicd avatar May 15 '25 17:05 tensorrt-cicd

PR_Github #5388 [ run ] triggered by Bot

tensorrt-cicd avatar May 15 '25 18:05 tensorrt-cicd

PR_Github #5388 [ run ] completed with state FAILURE /LLM/main/L0_MergeRequest_PR pipeline #3932 completed with status: 'FAILURE'

tensorrt-cicd avatar May 15 '25 21:05 tensorrt-cicd

/bot run --disable-fail-fast

PerkzZheng avatar May 16 '25 05:05 PerkzZheng

PR_Github #5454 [ run ] triggered by Bot

tensorrt-cicd avatar May 16 '25 05:05 tensorrt-cicd

/bot run --disable-fail-fast

PerkzZheng avatar May 16 '25 09:05 PerkzZheng

PR_Github #5501 [ run ] triggered by Bot

tensorrt-cicd avatar May 16 '25 10:05 tensorrt-cicd

PR_Github #5454 [ run ] completed with state ABORTED

tensorrt-cicd avatar May 16 '25 10:05 tensorrt-cicd

PR_Github #5501 [ run ] completed with state SUCCESS /LLM/main/L0_MergeRequest_PR pipeline #4008 completed with status: 'FAILURE'

tensorrt-cicd avatar May 16 '25 23:05 tensorrt-cicd

/bot run --only-multi-gpu-test

PerkzZheng avatar May 18 '25 08:05 PerkzZheng

PR_Github #5608 [ run ] triggered by Bot

tensorrt-cicd avatar May 18 '25 09:05 tensorrt-cicd