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
[CI/Build] Use CPU for mm processing test on CI (#27522)
Signed-off-by: Isotr0py <mozf@mail2.sysu.edu.cn>
Isotr0py committed
Oct 26, 2025 at 13:09 UTC
d63cd9ff1000a75cc88564e78317fd31d988dfa9
1 file changed
+4
-2
.buildkite/test-pipeline.yaml
+4
-2
@@ -733,14 +733,16 @@ steps:
733
commands:
734
- pytest -v -s models/language/pooling_mteb_test
735
736
-- label: Multi-Modal Processor Test # 44min
736
+- label: Multi-Modal Processor and Models Test (CPU) # 44min
737
timeout_in_minutes: 60
738
+ no_gpu: true
739
source_file_dependencies:
740
- vllm/
741
- tests/models/multimodal
742
commands:
743
- pip install git+https://github.com/TIGER-AI-Lab/Mantis.git
744
- pytest -v -s models/multimodal/processing
745
+ - pytest -v -s models/multimodal/test_mapping.py
746
747
- label: Multi-Modal Models Test (Standard) # 60min
748
timeout_in_minutes: 80
@@ -752,7 +754,7 @@ steps:
754
commands:
755
- pip install git+https://github.com/TIGER-AI-Lab/Mantis.git
756
- pip freeze | grep -E 'torch'
755
- - pytest -v -s models/multimodal -m core_model --ignore models/multimodal/generation/test_whisper.py --ignore models/multimodal/processing
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
758
- 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
759
760
- label: Multi-Modal Accuracy Eval (Small Models) # 50min