Matt Van Horn c31c50be51 fix(model): forward Qwen timeout and only set enable_thinking when true ...

Two bugs made local vLLM targets score acc=0.000: the router did not
forward 'timeout' to the Qwen backend (so runs used the 300s default),
and qwen_backend always injected chat_template_kwargs.enable_thinking,
which non-Qwen vLLM servers reject or answer with <think> output and no
<answer> tag. Forward timeout and only set the field when enabled.

Closes #28

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

2026-06-07 07:41:35 -07:00

..

__init__.py

fix(model): forward Qwen timeout and only set enable_thinking when true

2026-06-07 07:41:35 -07:00

azure_openai.py

Support Qwen chat as optimizer backend

2026-06-01 16:44:49 +08:00

backend_config.py

Support Qwen chat as optimizer backend

2026-06-01 16:44:49 +08:00

claude_backend.py

fix(claude): use --effort instead of deprecated --thinking flag

2026-05-30 11:24:13 +08:00

codex_backend.py

refactor: rename teacher/student to optimizer/target, remove best skills, fix slow update

2026-05-24 19:15:10 +00:00

codex_harness.py

refactor: rename teacher/student to optimizer/target, remove best skills, fix slow update

2026-05-24 19:15:10 +00:00

common.py

fix: use correct MiniMax endpoint, model name, and add .venv to gitignore

2026-05-31 05:27:50 +08:00

minimax_backend.py

fix: use correct MiniMax endpoint, model name, and add .venv to gitignore

2026-05-31 05:27:50 +08:00

qwen_backend.py

fix(model): forward Qwen timeout and only set enable_thinking when true

2026-06-07 07:41:35 -07:00

router.py

refactor: rename teacher/student to optimizer/target, remove best skills, fix slow update

2026-05-24 19:15:10 +00:00