@tduncan / Nvidia-Optimization-Stack-VLLM Mirror · read-only / commits / 7883c2bbe7
Fork 0

A high-throughput and memory-efficient inference and serving engine for LLMs

Mirrored from https://github.com/tduncan-lum/Nvidia-Optimization-Stack---VLLM.git

Synced Nov 16, 05:45 UTC

[benchmark] Make H100 benchmark optional (#10908)

Kevin H. Luu committed Dec 4, 2024 at 17:02 UTC 7883c2bbe7d0ab47160d205822f7b188a5a2771b
1 file changed +5
.buildkite/nightly-benchmarks/benchmark-pipeline.yaml
+5
@@ -65,10 +65,15 @@ steps:
65 - VLLM_USAGE_SOURCE
66 - HF_TOKEN
67
68 + - block: "Run H100 Benchmark"
69 + key: block-h100
70 + depends_on: ~
71 +
72 - label: "H100"
73 # skip: "use this flag to conditionally skip the benchmark step, useful for PR testing"
74 agents:
75 queue: H100
76 + depends_on: block-h100
77 plugins:
78 - docker#v5.12.0:
79 image: public.ecr.aws/q9t5s3a7/vllm-ci-postmerge-repo:$BUILDKITE_COMMIT