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 7, 00:15 UTC
Remove unused disable_fallback field (#36546)
Zhuohan Li committed
Mar 9, 2026 at 20:56 UTC
04b67d8f62cab3a1832df5c6ed840f8a6afccaf9
2 files changed
-3
vllm/config/structured_outputs.py
-2
@@ -23,8 +23,6 @@ class StructuredOutputsConfig:
23
regex, etc) by default. With "auto", we will make opinionated choices
24
based on request contents and what the backend libraries currently support,
25
so the behavior is subject to change in each release."""
26
- disable_fallback: bool = False
27
- """If `True`, vLLM will not fallback to a different backend on error."""
26
disable_any_whitespace: bool = False
27
"""If `True`, json output will always be compact without any whitespace.
28
If `False`, the model may generate whitespace between JSON fields,
vllm/sampling_params.py
-1
@@ -41,7 +41,6 @@ class StructuredOutputsParams:
41
grammar: str | None = None
42
json_object: bool | None = None
43
# These are other options that can be set.
44
- disable_fallback: bool = False
44
disable_any_whitespace: bool = False
45
disable_additional_properties: bool = False
46
whitespace_pattern: str | None = None