[NVBUG-5304516/5319741]Qwen2.5VL FP8 support

[DylanChen-NV]

[NVBUG-5304516/5319741]Qwen2.5VL FP8 support

Description

1. Add FP8 cuda core GEMM in torch flow to optimize performance for Ada. The e2e performance of BS=1 generation phase on L20 is BF16 : FP8 cublas : FP8 cudacore = 1 : 1.07 : 1.85.
2. Release the multi-modal feature in requests in time to avoid GPU memory leak (NVBUG-5304516).
3. Fix the issue that using cuda graph leads to abnormal outputs (NVBUG-5319741).
4. Use multi_threading in request preprocessing to overlap vision model inference and IO of requests. Speedup the e2e performance by 35% for 720p image + 1k LLM input + 120 LLM output. (Reverted, because a better way to solve is [TRTLLM-4958][feat] improve multimodal workflow with Vision + LLM inflight batching #4673)

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.

Details

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.

[DylanChen-NV]

/bot run

[tensorrt-cicd]

PR_Github #8118 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #8118 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #5885 completed with status: 'FAILURE'

[DylanChen-NV]

/bot run

[tensorrt-cicd]

PR_Github #8199 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #8199 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #5945 completed with status: 'FAILURE'

[DylanChen-NV]

/bot run

[tensorrt-cicd]

PR_Github #8249 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #8249 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #5976 completed with status: 'FAILURE'

[DylanChen-NV]

/bot run --stage-list "B200_PCIe-PyTorch-1"

[tensorrt-cicd]

PR_Github #8383 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #8383 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #6075 (Partly Tested) completed with status: 'SUCCESS'

[DylanChen-NV]

/bot run

[tensorrt-cicd]

PR_Github #8608 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #8608 [ run ] completed with state FAILURE
/LLM/main/L0_MergeRequest_PR pipeline #6247 completed with status: 'FAILURE'

[tensorrt-cicd]

PR_Github #10908 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #10908 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8062 completed with status: 'SUCCESS'
Pipeline passed with automatic retried tests. Check the rerun report for details.

[symphonylyh]

lgtm once Chang's comment is addressed #5029 (comment)

[DylanChen-NV]

/bot run

[tensorrt-cicd]

PR_Github #11316 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11316 [ run ] completed with state FAILURE
/LLM/main/L0_MergeRequest_PR pipeline #8369 completed with status: 'FAILURE'

[DylanChen-NV]

/bot run

[tensorrt-cicd]

PR_Github #11334 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11334 [ run ] completed with state FAILURE
/LLM/main/L0_MergeRequest_PR pipeline #8385 completed with status: 'FAILURE'

[DylanChen-NV]

/bot run

[tensorrt-cicd]

PR_Github #11377 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11377 [ run ] completed with state FAILURE
/LLM/main/L0_MergeRequest_PR pipeline #8415 completed with status: 'FAILURE'

[DylanChen-NV]

/bot run

[tensorrt-cicd]

PR_Github #11393 [ run ] triggered by Bot

[tensorrt-cicd]

PR_Github #11393 [ run ] completed with state SUCCESS
/LLM/main/L0_MergeRequest_PR pipeline #8424 completed with status: 'SUCCESS'