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 15, 08:30 UTC
[CI/Build] Split pooling and generation extended language models tests in CI (#18705)
Signed-off-by: Isotr0py <2037008807@qq.com>
Isotr0py committed
May 26, 2025 at 19:00 UTC
08777500295422577fb7d88b2eb798a3d56af3ee
1 file changed
+12
-3
.buildkite/test-pipeline.yaml
+12
-3
@@ -486,16 +486,25 @@ steps:
486
- pip freeze | grep -E 'torch'
487
- pytest -v -s models/language -m core_model
488
489
-- label: Language Models Test (Extended)
489
+- label: Language Models Test (Extended Generation) # 1hr20min
490
mirror_hardwares: [amdexperimental]
491
optional: true
492
source_file_dependencies:
493
- vllm/
494
- - tests/models/language
494
+ - tests/models/language/generation
495
commands:
496
# Install causal-conv1d for plamo2 models here, as it is not compatible with pip-compile.
497
- pip install 'git+https://github.com/Dao-AILab/causal-conv1d@v1.5.0.post8'
498
- - pytest -v -s models/language -m 'not core_model'
498
+ - pytest -v -s models/language/generation -m 'not core_model'
499
+
500
+- label: Language Models Test (Extended Pooling) # 36min
501
+ mirror_hardwares: [amdexperimental]
502
+ optional: true
503
+ source_file_dependencies:
504
+ - vllm/
505
+ - tests/models/language/pooling
506
+ commands:
507
+ - pytest -v -s models/language/pooling -m 'not core_model'
508
509
- label: Multi-Modal Models Test (Standard)
510
mirror_hardwares: [amdexperimental]