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 5, 01:00 UTC
[Misc][UX] Map mimo reasoning and tooling parsers (#40089)
Signed-off-by: Roger Wang <hey@rogerw.io> Co-authored-by: Chauncey <chaunceyjiang@gmail.com>
Roger Wang committed
Apr 17, 2026 at 09:49 UTC
512765d52d743ebaefb7705ff775226d6a7fcc7a
2 files changed
+8
vllm/reasoning/__init__.py
+4
@@ -60,6 +60,10 @@ _REASONING_PARSERS_TO_REGISTER = {
60
"kimi_k2_reasoning_parser",
61
"KimiK2ReasoningParser",
62
),
63
+ "mimo": (
64
+ "qwen3_reasoning_parser",
65
+ "Qwen3ReasoningParser",
66
+ ),
67
"minimax_m2": (
68
"minimax_m2_reasoning_parser",
69
"MiniMaxM2ReasoningParser",
vllm/tool_parsers/__init__.py
+4
@@ -94,6 +94,10 @@ _TOOL_PARSERS_TO_REGISTER = {
94
"longcat_tool_parser",
95
"LongcatFlashToolParser",
96
),
97
+ "mimo": (
98
+ "qwen3xml_tool_parser",
99
+ "Qwen3XMLToolParser",
100
+ ),
101
"minimax_m2": (
102
"minimax_m2_tool_parser",
103
"MinimaxM2ToolParser",