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 11, 15:00 UTC
[ROCm][CI] Use mi325_4 agent pool for V1 e2e tests (#30526)
Signed-off-by: Andreas Karatzas <akaratza@amd.com>
Andreas Karatzas committed
Dec 11, 2025 at 19:37 UTC
b5945d49c08b66658110fa1c63e55fde66fcfad7
1 file changed
+3
-3
.buildkite/test-amd.yaml
+3
-3
@@ -326,10 +326,10 @@ steps:
326
commands:
327
- pytest -v -s engine test_sequence.py test_config.py test_logger.py test_vllm_port.py
328
329
-- label: V1 Test e2e + engine # 30min
330
- timeout_in_minutes: 45
329
+- label: V1 Test e2e + engine # 65min
330
+ timeout_in_minutes: 90
331
mirror_hardwares: [amdexperimental]
332
- agent_pool: mi325_1
332
+ agent_pool: mi325_4
333
# grade: Blocking
334
source_file_dependencies:
335
- vllm/