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 14, 05:30 UTC
[CI] Disable flaky structured output test from CI (#24366)
Signed-off-by: Roger Wang <hey@rogerw.io>
Roger Wang committed
Sep 6, 2025 at 06:31 UTC
b121ca22ad9a648513af098052915dd0afb1dd47
1 file changed
+3
-2
tests/v1/entrypoints/llm/test_struct_output_generate.py
+3
-2
@@ -41,8 +41,9 @@ EAGLE_SPEC_CONFIG = {
41
PARAMS_MODELS_BACKENDS_TOKENIZER_MODE = [
42
("mistralai/Ministral-8B-Instruct-2410", "xgrammar", "auto", None),
43
("mistralai/Ministral-8B-Instruct-2410", "guidance", "auto", None),
44
- ("mistralai/Ministral-8B-Instruct-2410", "lm-format-enforcer", "auto",
45
- None),
44
+ #FIXME: This test is flaky on CI thus disabled
45
+ #("mistralai/Ministral-8B-Instruct-2410", "lm-format-enforcer", "auto",
46
+ # None),
47
("mistralai/Ministral-8B-Instruct-2410", "xgrammar", "mistral", None),
48
("Qwen/Qwen2.5-1.5B-Instruct", "xgrammar", "auto", None),
49
("Qwen/Qwen2.5-1.5B-Instruct", "lm-format-enforcer", "auto", None),