Files
project_6/vllm
project6 bf6ceb0b12 fix(critical): disable thinking for tool_call requests — fixes d03_tool_call + d05/t5 FAIL
Root cause: Model spends all tokens in <think>...</think> instead of emitting
<tool_call> XML. Competitor Sub168 completes d03 in 2.12s; we took 49s and FAIL.

Fix: When tool_choice != 'none' and tools present, inject enable_thinking=False
into chat_template_kwargs before calling apply_hf_chat_template().

Also handles OpenAI-style thinking field and adds competitive analysis doc.
2026-08-07 08:45:22 +00:00
..