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 3, 21:30 UTC
[CI] Migrate more B200 jobs to b200-k8s queue (#42356)
Signed-off-by: khluu <khluu000@gmail.com>
Kevin H. Luu committed
May 12, 2026 at 00:38 UTC
f69644caf8532df1a6e610d2bfbdc86d9a334186
3 files changed
+4
-4
.buildkite/test_areas/kernels.yaml
+1
-1
@@ -242,7 +242,7 @@ steps:
242
- label: Kernels FusedMoE Layer Test (2 B200s)
243
key: kernels-fusedmoe-layer-test-2-b200s
244
timeout_in_minutes: 90
245
- device: b200
245
+ device: b200-k8s
246
num_devices: 2
247
source_file_dependencies:
248
- csrc/quantization/cutlass_w8a8/moe/
.buildkite/test_areas/lm_eval.yaml
+2
-2
@@ -92,7 +92,7 @@ steps:
92
93
- label: MoE Refactor Integration Test (B200 DP - TEMPORARY)
94
key: moe-refactor-integration-test-b200-dp-temporary
95
- device: b200
95
+ device: b200-k8s
96
optional: true
97
num_devices: 2
98
commands:
@@ -127,7 +127,7 @@ steps:
127
- label: GPQA Eval (GPT-OSS) (B200)
128
key: gpqa-eval-gpt-oss-b200
129
timeout_in_minutes: 120
130
- device: b200
130
+ device: b200-k8s
131
optional: true
132
num_devices: 2
133
source_file_dependencies:
.buildkite/test_areas/spec_decode.yaml
+1
-1
@@ -100,7 +100,7 @@ steps:
100
101
- label: Spec Decode MTP hybrid (B200)
102
timeout_in_minutes: 30
103
- device: b200
103
+ device: b200-k8s
104
optional: true
105
source_file_dependencies:
106
- vllm/v1/spec_decode/