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 13, 05:45 UTC
[BugFix][CI/Build][ROCM] Fix import error and apply assert in appropriate case in test_struct_output_generate (#28311)
Signed-off-by: Randall Smith <ransmith@amd.com> Co-authored-by: Randall Smith <ransmith@amd.com>
rasmith committed
Nov 14, 2025 at 00:41 UTC
93103575ce0480f36fc1a3603eb51d9a89f38a00
1 file changed
+8
-3
tests/v1/entrypoints/llm/test_struct_output_generate.py
+8
-3
@@ -677,9 +677,14 @@ def test_structured_output_with_reasoning_matrices(
677
reasoning, content = run_reasoning_extraction(reasoner, [generated_text])
678
print(f"Prompt: {prompt!r}\nReasoning: {reasoning!r}\nContent: {content!r}")
679
680
- assert content is not None and reasoning is not None
681
- output_json = json.loads(content)
682
- jsonschema.validate(instance=output_json, schema=reasoning_schema)
680
+ if "Qwen3" in model_name:
681
+ assert content is not None
682
+
683
+ assert reasoning is not None
684
+
685
+ if content is not None:
686
+ output_json = json.loads(content)
687
+ jsonschema.validate(instance=output_json, schema=reasoning_schema)
688
689
690
@pytest.mark.skip_global_cleanup