diff --git a/.buildkite/test-amd.yaml b/.buildkite/test-amd.yaml index 789ca54260..17b0fa408e 100644 --- a/.buildkite/test-amd.yaml +++ b/.buildkite/test-amd.yaml @@ -1341,21 +1341,6 @@ steps: - export VLLM_WORKER_MULTIPROC_METHOD=spawn - pytest -v -s entrypoints/multimodal -- label: ":amd: (MI300) Entrypoints Integration (Pooling)" # TBD - timeout_in_minutes: 65 - mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] - dind: false - agent_pool: mi300_1 - optional: true - fast_check: true - working_dir: "/vllm-workspace/tests" - source_file_dependencies: - - vllm/ - - tests/entrypoints/pooling - commands: - - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s entrypoints/pooling - - label: ":amd: (MI300) OpenAI API Correctness" # TBD timeout_in_minutes: 30 mirror_hardwares: [amdexperimental, amdproduction, amdgfx942nightly, amdmi300] @@ -3615,12 +3600,13 @@ steps: - export VLLM_WORKER_MULTIPROC_METHOD=spawn - pytest -v -s entrypoints/multimodal -- label: ":amd: (MI355) Entrypoints Integration (Pooling)" # TBD +- label: ":amd: (MI355) Entrypoints Integration (Pooling) Shard %N" # TBD timeout_in_minutes: 65 mirror_hardwares: [amdexperimental, amdproduction, amdgfx950nightly, amdmi355] dind: false agent_pool: mi355_1 fast_check: true + parallelism: 4 working_dir: "/vllm-workspace/tests" source_file_dependencies: - vllm/ @@ -3628,7 +3614,7 @@ steps: optional: true commands: - export VLLM_WORKER_MULTIPROC_METHOD=spawn - - pytest -v -s entrypoints/pooling + - pytest -v -s entrypoints/pooling --num-shards=$$BUILDKITE_PARALLEL_JOB_COUNT --shard-id=$$BUILDKITE_PARALLEL_JOB #----------------------------------------------------------- mi355 ยท evals -----------------------------------------------------------#