[CI] Split V1 e2e + engine (1 GPU) into separate jobs (#36945)

Co-authored-by: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
Kevin H. Luu
2026-03-13 14:16:02 -07:00
committed by GitHub
co-authored by Claude Opus 4.6
parent 0005d2a3c9
commit f1816fb192
18 changed files with 81 additions and 39 deletions
View File
@@ -14,8 +14,8 @@ from vllm.platforms import current_platform
from vllm.sampling_params import StructuredOutputsParams
from vllm.v1.metrics.reader import Metric
from ...conftest import VllmRunner
from ...models.utils import check_outputs_equal
from ....conftest import VllmRunner
from ....models.utils import check_outputs_equal
MODEL = "Qwen/Qwen3-0.6B"
MTP_MODEL = "meta-llama/Llama-3.2-1B-Instruct"
@@ -5,7 +5,7 @@ import pytest
from vllm import LLM, SamplingParams
from ...utils import create_new_process_for_each_test
from ....utils import create_new_process_for_each_test
@create_new_process_for_each_test()
@@ -7,7 +7,7 @@ import pytest
from vllm import LLM, SamplingParams
from vllm.platforms import current_platform
from ...utils import check_answers, prep_prompts
from ....utils import check_answers, prep_prompts
@dataclass
@@ -9,7 +9,7 @@ from vllm import LLM, SamplingParams
from vllm.config import CompilationConfig, CompilationMode
from vllm.platforms import current_platform
from ...utils import check_answers, fork_new_process_for_each_test, prep_prompts
from ....utils import check_answers, fork_new_process_for_each_test, prep_prompts
# global seed
SEED = 42
@@ -18,7 +18,7 @@ SEED = 42
@pytest.fixture
def test_prompts():
"""
Adapted from tests/v1/e2e/test_spec_decode.py
Adapted from tests/v1/e2e/spec_decode/test_spec_decode.py
"""
prompt_types = ["repeat", "sentence"]
# Setting higher num prompts increases the chance of numerics mismatch
@@ -497,6 +497,6 @@ if __name__ == "__main__":
Usage:
cd vllm/
python -m pytest tests/v1/e2e/test_min_tokens.py -v
python -m pytest tests/v1/e2e/general/test_min_tokens.py -v
"""
pytest.main([__file__, "-v"])