7777 uv pip install --no-upgrade --system pytest pytest-asyncio
7878 uv pip install --no-upgrade --system -e tests/vllm_test_utils
7979 uv pip install --no-upgrade --system hf_transfer
80+ uv pip install --system model-hosting-container-standards==0.1.9
8081 mkdir src
8182 mv vllm src/vllm
8283 '
8889 nvidia-smi
8990
9091 pip list | grep model-hosting-container-standards
91- SAGEMAKER_CONTAINER_LOG_LEVEL=20 vllm serve TinyLlama/TinyLlama-1.1B-Chat-v1.0 --dtype auto
92- SAGEMAKER_CONTAINER_LOG_LEVEL=INFO vllm serve TinyLlama/TinyLlama-1.1B-Chat-v1.0 --dtype auto
92+
93+ SAGEMAKER_CONTAINER_LOG_LEVEL=20 vllm serve TinyLlama/TinyLlama-1.1B-Chat-v1.0 --dtype auto &
94+ VLLM_PID=$!
95+ sleep 60
96+ kill $VLLM_PID
97+
98+ SAGEMAKER_CONTAINER_LOG_LEVEL=INFO vllm serve TinyLlama/TinyLlama-1.1B-Chat-v1.0 --dtype auto &
99+ VLLM_PID=$!
100+ sleep 60
101+ kill $VLLM_PID
93102
94103 # Test LoRA adapter loading/unloading via SageMaker endpoints
95104 pytest tests/entrypoints/sagemaker/test_sagemaker_lora_adapters.py -v
@@ -210,6 +219,7 @@ jobs:
210219 uv pip install --no-upgrade --system pytest pytest-asyncio
211220 uv pip install --no-upgrade --system -e tests/vllm_test_utils
212221 uv pip install --no-upgrade --system hf_transfer
222+ uv pip install --system model-hosting-container-standards==0.1.9
213223 mkdir src
214224 mv vllm src/vllm
215225 '
@@ -274,6 +284,7 @@ jobs:
274284 uv pip install --no-upgrade --system pytest pytest-asyncio
275285 uv pip install --no-upgrade --system -e tests/vllm_test_utils
276286 uv pip install --no-upgrade --system hf_transfer
287+ uv pip install --system model-hosting-container-standards==0.1.9
277288 mkdir src
278289 mv vllm src/vllm
279290 '
@@ -356,6 +367,7 @@ jobs:
356367 uv pip install --no-upgrade --system pytest pytest-asyncio
357368 uv pip install --no-upgrade --system -e tests/vllm_test_utils
358369 uv pip install --no-upgrade --system hf_transfer
370+ uv pip install --system model-hosting-container-standards==0.1.9
359371 mkdir src
360372 mv vllm src/vllm
361373 '
@@ -367,8 +379,16 @@ jobs:
367379 nvidia-smi
368380
369381 pip list | grep model-hosting-container-standards
370- SAGEMAKER_CONTAINER_LOG_LEVEL=20 vllm serve TinyLlama/TinyLlama-1.1B-Chat-v1.0 --dtype auto
371- SAGEMAKER_CONTAINER_LOG_LEVEL=INFO vllm serve TinyLlama/TinyLlama-1.1B-Chat-v1.0 --dtype auto
382+
383+ SAGEMAKER_CONTAINER_LOG_LEVEL=20 vllm serve TinyLlama/TinyLlama-1.1B-Chat-v1.0 --dtype auto &
384+ VLLM_PID=$!
385+ sleep 60
386+ kill $VLLM_PID
387+
388+ SAGEMAKER_CONTAINER_LOG_LEVEL=INFO vllm serve TinyLlama/TinyLlama-1.1B-Chat-v1.0 --dtype auto &
389+ VLLM_PID=$!
390+ sleep 60
391+ kill $VLLM_PID
372392
373393 # Test LoRA adapter loading/unloading via SageMaker endpoints
374394 pytest tests/entrypoints/sagemaker/test_sagemaker_lora_adapters.py -v
@@ -484,6 +504,7 @@ jobs:
484504 uv pip install --no-upgrade --system pytest pytest-asyncio
485505 uv pip install --no-upgrade --system -e tests/vllm_test_utils
486506 uv pip install --no-upgrade --system hf_transfer
507+ uv pip install --system model-hosting-container-standards==0.1.9
487508 mkdir src
488509 mv vllm src/vllm
489510 '
@@ -548,6 +569,7 @@ jobs:
548569 uv pip install --no-upgrade --system pytest pytest-asyncio
549570 uv pip install --no-upgrade --system -e tests/vllm_test_utils
550571 uv pip install --no-upgrade --system hf_transfer
572+ uv pip install --system model-hosting-container-standards==0.1.9
551573 mkdir src
552574 mv vllm src/vllm
553575 '
@@ -611,6 +633,7 @@ jobs:
611633 uv pip install --no-upgrade --system pytest pytest-asyncio
612634 uv pip install --no-upgrade --system -e tests/vllm_test_utils
613635 uv pip install --no-upgrade --system hf_transfer
636+ uv pip install --system model-hosting-container-standards==0.1.9
614637 mkdir src
615638 mv vllm src/vllm
616639 '
0 commit comments