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 12, 19:45 UTC
[CI/Build] Remove skip global cleanup in test_struct_output_generate.py (#29022)
Signed-off-by: Randall Smith <ransmith@amd.com> Co-authored-by: Randall Smith <ransmith@amd.com>
rasmith committed
Nov 20, 2025 at 15:44 UTC
c7a29d2c8d07ce6188d0c4bb19df6fd1d0e9bc74
1 file changed
-4
tests/v1/entrypoints/llm/test_struct_output_generate.py
-4
@@ -121,7 +121,6 @@ def test_guided_decoding_deprecated():
121
assert sp1.structured_outputs == guided_decoding
122
123
124
-@pytest.mark.skip_global_cleanup
124
@pytest.mark.parametrize(
125
"model_name, backend, tokenizer_mode, speculative_config",
126
PARAMS_MODELS_BACKENDS_TOKENIZER_MODE,
@@ -626,7 +625,6 @@ Make the response as short as possible.
625
)
626
627
629
-@pytest.mark.skip_global_cleanup
628
@pytest.mark.parametrize(
629
"model_name, backend, tokenizer_mode, reasoning_parser, speculative_config", # noqa: E501
630
[
@@ -711,7 +709,6 @@ def test_structured_output_with_reasoning_matrices(
709
jsonschema.validate(instance=output_json, schema=reasoning_schema)
710
711
714
-@pytest.mark.skip_global_cleanup
712
@pytest.mark.parametrize("model_name, tokenizer_mode", PARAMS_MODELS_TOKENIZER_MODE)
713
def test_structured_output_auto_mode(
714
unsupported_json_schema: dict[str, Any],
@@ -758,7 +755,6 @@ def test_structured_output_auto_mode(
755
assert isinstance(parsed_json, dict)
756
757
761
-@pytest.mark.skip_global_cleanup
758
def test_guidance_no_additional_properties():
759
llm = LLM(
760
model="Qwen/Qwen2.5-1.5B-Instruct",