Andrew Peabody 7708ffa986 fix(llm): support Gemini flash-lite & reasoning models, strip extra_body on 400 retry (#4342) ...

- Define _NO_TEMPERATURE_MODEL_KEYWORDS constant and omit temperature=0 on initial request
- Omit thinkingConfig in _thinking_extra_body for flash-lite models
- Unconditionally strip rejected sampling params and extra_body thinkingConfig on BadRequestError (handling Gemini 400 INVALID_ARGUMENT without named details)
- Add comprehensive unit tests in test_llm_client_gemini.py

2026-09-02 10:45:39 +02:00

..

__init__.py

LLM / AI Change detection rules and Summaries

2026-04-24 15:21:09 +02:00

bm25_trim.py

LLM / AI Change detection rules and Summaries

2026-04-24 15:21:09 +02:00

client.py

fix(llm): support Gemini flash-lite & reasoning models, strip extra_body on 400 retry (#4342)

2026-09-02 10:45:39 +02:00

evaluator.py

fix(llm): support Gemini flash-lite & reasoning models, strip extra_body on 400 retry (#4342)

2026-09-02 10:45:39 +02:00

prompt_builder.py

fix(llm): string boolean parsing, reasoning model <think> stripping, and timestamp diff annotation (#4340)

2026-09-01 11:52:40 +02:00

response_parser.py

fix(llm): handle reasoning blocks that aren't a well-formed <think>...</think> pair (#4349)

2026-09-01 12:42:16 +02:00

ui_strings.py

LLM / AI Change detection rules and Summaries

2026-04-24 15:21:09 +02:00