diff --git a/gpttype_adapter.cpp b/gpttype_adapter.cpp index 9840e5c4a..6098f6eab 100644 --- a/gpttype_adapter.cpp +++ b/gpttype_adapter.cpp @@ -5859,6 +5859,11 @@ generation_outputs gpttype_generate(const generation_inputs inputs) end = "<|END_THINKING|>"; budget_exceeded = "\n(Reasoning budget exceeded)\nTime to respond now.\n<|END_THINKING|>"; break; + case llm_arch::LLM_ARCH_MISTRAL3: + start = "[THINK]"; + end = "[/THINK]"; + budget_exceeded = "\n(Reasoning budget exceeded)\nTime to respond now.\n[/THINK]"; + break; default: break; } diff --git a/koboldcpp.py b/koboldcpp.py index 9ac7aba1c..fd3f967a1 100644 --- a/koboldcpp.py +++ b/koboldcpp.py @@ -186,7 +186,8 @@ thinkformats = [{"start":"<|channel|>analysis<|message|>","end":"<|start|>assist {"start":"","end":""}, {"start":"","end":""}, {"start":"<|START_THINKING|>","end":"<|END_THINKING|>"}, - {"start":"<|channel>thought","end":""}] + {"start":"<|channel>thought","end":""}, + {"start":"[THINK]","end":"[/THINK]"}] tool_call_pairs = [ #third element is optional str to match in chat template before we use this pair, fourth element is whether its stream-handleable ("", "", None, True), #qwen, glm ("", "", None, True), #seed oss