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 8, 03:45 UTC
[CI/Build] Remove use of `skip_v1` (#34699)
Signed-off-by: DarkLight1337 <tlleungac@connect.ust.hk>
Cyrus Leung committed
Feb 18, 2026 at 12:19 UTC
30ebe0dc3c24e016b5e6f2f4532939867719097e
6 files changed
+9
-5
.buildkite/test-amd.yaml
+4
-2
@@ -55,9 +55,11 @@ steps:
55
grade: Blocking
56
source_file_dependencies:
57
- vllm/
58
+ - tests/detokenizer
59
- tests/multimodal
60
- tests/utils_
61
commands:
62
+ - pytest -v -s detokenizer
63
- pytest -v -s -m 'not cpu_test' multimodal
64
- pytest -v -s utils_
65
@@ -547,7 +549,7 @@ steps:
549
- tests/samplers
550
- tests/conftest.py
551
commands:
550
- - pytest -v -s -m 'not skip_v1' samplers
552
+ - pytest -v -s -m samplers
553
554
- label: LoRA Test %N # 20min each
555
timeout_in_minutes: 30
@@ -2213,7 +2215,7 @@ steps:
2215
- tests/samplers
2216
- tests/conftest.py
2217
commands:
2216
- - pytest -v -s -m 'not skip_v1' samplers
2218
+ - pytest -v -s -m samplers
2219
2220
- label: LoRA Test %N # 20min each
2221
timeout_in_minutes: 30
.buildkite/test-pipeline.yaml
+2
@@ -51,9 +51,11 @@ steps:
51
mirror_hardwares: [amdexperimental]
52
source_file_dependencies:
53
- vllm/
54
+ - tests/detokenizer
55
- tests/multimodal
56
- tests/utils_
57
commands:
58
+ - pytest -v -s detokenizer
59
- pytest -v -s -m 'not cpu_test' multimodal
60
- pytest -v -s utils_
61
.buildkite/test_areas/misc.yaml
+2
@@ -108,9 +108,11 @@ steps:
108
timeout_in_minutes: 50
109
source_file_dependencies:
110
- vllm/
111
+ - tests/detokenizer
112
- tests/multimodal
113
- tests/utils_
114
commands:
115
+ - pytest -v -s detokenizer
116
- pytest -v -s -m 'not cpu_test' multimodal
117
- pytest -v -s utils_
118
.buildkite/test_areas/samplers.yaml
+1
-1
@@ -18,4 +18,4 @@ steps:
18
depends_on:
19
- image-build-amd
20
commands:
21
- - pytest -v -s -m 'not skip_v1' samplers
21
+ - pytest -v -s -m samplers
pyproject.toml
-1
@@ -113,7 +113,6 @@ markers = [
113
"cpu_test: mark test as CPU-only test",
114
"split: run this test as part of a split",
115
"distributed: run this test only in distributed GPU tests",
116
- "skip_v1: do not run this test with v1",
116
"optional: optional tests that are automatically skipped, include --optional to run them",
117
]
118
tests/detokenizer/test_disable_detokenization.py
-1
@@ -7,7 +7,6 @@ from vllm.entrypoints.llm import LLM
7
from vllm.sampling_params import SamplingParams
8
9
10
-@pytest.mark.skip_v1
10
@pytest.mark.parametrize("model", ["distilbert/distilgpt2"])
11
def test_computed_prefix_blocks(model: str):
12
# This test checks if the engine generates completions both with and