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] Skip lm-format-enforcer tests in test_struct_output_generate.py for now (#29021)
Signed-off-by: Randall Smith <ransmith@amd.com> Co-authored-by: Randall Smith <ransmith@amd.com>
rasmith committed
Nov 20, 2025 at 15:35 UTC
8237ab8a2bed14bec5cafbec75033c8e1d54d852
1 file changed
+26
-2
tests/v1/entrypoints/llm/test_struct_output_generate.py
+26
-2
@@ -47,10 +47,34 @@ EAGLE_SPEC_CONFIG = {
47
PARAMS_MODELS_BACKENDS_TOKENIZER_MODE = [
48
("mistralai/Ministral-8B-Instruct-2410", "xgrammar", "auto", None),
49
("mistralai/Ministral-8B-Instruct-2410", "guidance", "auto", None),
50
- ("mistralai/Ministral-8B-Instruct-2410", "lm-format-enforcer", "auto", None),
50
+ pytest.param(
51
+ "mistralai/Ministral-8B-Instruct-2410",
52
+ "lm-format-enforcer",
53
+ "auto",
54
+ None,
55
+ marks=pytest.mark.skip(
56
+ reason=(
57
+ "Flaky: lm-format-enforcer intermittently returns"
58
+ "incomplete JSON."
59
+ "See https://github.com/noamgat/lm-format-enforcer/issues/169"
60
+ )
61
+ ),
62
+ ),
63
("mistralai/Ministral-8B-Instruct-2410", "xgrammar", "mistral", None),
64
("Qwen/Qwen2.5-1.5B-Instruct", "xgrammar", "auto", None),
53
- ("Qwen/Qwen2.5-1.5B-Instruct", "lm-format-enforcer", "auto", None),
65
+ pytest.param(
66
+ "Qwen/Qwen2.5-1.5B-Instruct",
67
+ "lm-format-enforcer",
68
+ "auto",
69
+ None,
70
+ marks=pytest.mark.skip(
71
+ reason=(
72
+ "Flaky: lm-format-enforcer intermittently returns"
73
+ "incomplete JSON."
74
+ "See https://github.com/noamgat/lm-format-enforcer/issues/169"
75
+ )
76
+ ),
77
+ ),
78
# FIXME: This tests are flaky on CI thus disabled. Tracking in Issue #24402
79
# ("mistralai/Ministral-8B-Instruct-2410", "outlines", "auto", None),
80
# ("mistralai/Ministral-8B-Instruct-2410", "outlines", "mistral", None),