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 14, 05:30 UTC
Revert "[CI/Build] Use CPU for mm processing test on CI (#27522)" (#27531)
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Cyrus Leung committed
Oct 26, 2025 at 19:44 UTC
c7abff299021a9562dc8b50a82509dd2af28eddb
1 file changed
+2
-4
.buildkite/test-pipeline.yaml
+2
-4
@@ -733,16 +733,14 @@ steps:
733
commands:
734
- pytest -v -s models/language/pooling_mteb_test
735
736
-- label: Multi-Modal Processor and Models Test (CPU) # 44min
736
+- label: Multi-Modal Processor Test # 44min
737
timeout_in_minutes: 60
738
- no_gpu: true
738
source_file_dependencies:
739
- vllm/
740
- tests/models/multimodal
741
commands:
742
- pip install git+https://github.com/TIGER-AI-Lab/Mantis.git
743
- pytest -v -s models/multimodal/processing
745
- - pytest -v -s models/multimodal/test_mapping.py
744
745
- label: Multi-Modal Models Test (Standard) # 60min
746
timeout_in_minutes: 80
@@ -754,7 +752,7 @@ steps:
752
commands:
753
- pip install git+https://github.com/TIGER-AI-Lab/Mantis.git
754
- pip freeze | grep -E 'torch'
757
- - pytest -v -s models/multimodal -m core_model --ignore models/multimodal/test_mapping.py --ignore models/multimodal/generation/test_whisper.py --ignore models/multimodal/processing
755
+ - pytest -v -s models/multimodal -m core_model --ignore models/multimodal/generation/test_whisper.py --ignore models/multimodal/processing
756
- cd .. && VLLM_WORKER_MULTIPROC_METHOD=spawn pytest -v -s tests/models/multimodal/generation/test_whisper.py -m core_model # Otherwise, mp_method="spawn" doesn't work
757
758
- label: Multi-Modal Accuracy Eval (Small Models) # 50min