curo/kite - kite - Gitea: Git with a cup of tea

471 Commits 1 Branch 1 Tag

2 Commits

Include renames

Author	SHA1	Message	Date
curo1305andClaude Sonnet 4.6	ac2dded35b	fix(celery+lmstudio): PYTHONPATH for forked workers + qwen3.5-9b thinking model support Three root causes fixed: 1. docker-compose.yml — celery-worker and celery-beat missing PYTHONPATH=/app ForkPoolWorker processes inherit sys.path with '' (CWD), but fork can change CWD so lazy imports like `from db.session import ...` raised ModuleNotFoundError. Adding PYTHONPATH=/app ensures /app is always explicit in the path. 2. provider_config.py — lmstudio SUPPORTS_JSON_MODE set to False LM Studio only accepts response_format type 'json_schema' or 'text', not 'json_object'. GenericOpenAIProvider now omits response_format for lmstudio and relies on parse_classification() fallback (same path as Gemini). 3. generic_openai_provider.py — max_tokens raised 1024→4096 (classify) / 256→1024 (suggest) Qwen3 thinking models (like qwen/qwen3.5-9b) consume reasoning tokens from the same budget. With max_tokens=1024 the model exhausts the budget in the thinking trace, leaving content=''. 4096 gives room for both thinking and the JSON output. Also updates lmstudio PROVIDER_DEFAULTS model to qwen/qwen3.5-9b (confirmed loaded in LM Studio) and context_chars to 32_000 (Qwen3 context window). Tested: classify(invoice_text, ['Finance','Legal','HR']) → topics=['Finance'] ✓ Backend tests: 376 passed ✓ Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>	2026-06-05 00:00:26 +02:00
curo1305	02bcbb9143	feat(07-02): singleton OpenAIProvider + GenericOpenAIProvider + MAX_AI_CHARS removal - openai_provider.py: singleton self._client=AsyncOpenAI(...) in __init__ (D-07), _truncate() 60/40 smart truncation (D-13), removed MAX_AI_CHARS constant, changed __init__ signature to include context_chars, removed _client() method - generic_openai_provider.py: new class, subclasses OpenAIProvider, conditional response_format={"type":"json_object"} on supports_json_mode flag (D-01/D-02), imports parse_classification + parse_suggestions from ai.utils (D-02 contract) - ollama_provider.py: added context_chars kwarg with default 8000, passes through - lmstudio_provider.py: added context_chars kwarg with default 8000, passes through - classifier.py: removed MAX_AI_CHARS constant and text[:MAX_AI_CHARS] slices; truncation now handled inside each provider via _truncate() - requirements.txt: bumped anthropic floor to >=0.95.0 (D-03 output_config support)	2026-06-04 18:58:19 +02:00

Author

SHA1

Message

Date

curo1305andClaude Sonnet 4.6

ac2dded35b

fix(celery+lmstudio): PYTHONPATH for forked workers + qwen3.5-9b thinking model support

Three root causes fixed:

1. docker-compose.yml — celery-worker and celery-beat missing PYTHONPATH=/app
   ForkPoolWorker processes inherit sys.path with '' (CWD), but fork can change
   CWD so lazy imports like `from db.session import ...` raised ModuleNotFoundError.
   Adding PYTHONPATH=/app ensures /app is always explicit in the path.

2. provider_config.py — lmstudio SUPPORTS_JSON_MODE set to False
   LM Studio only accepts response_format type 'json_schema' or 'text', not
   'json_object'. GenericOpenAIProvider now omits response_format for lmstudio
   and relies on parse_classification() fallback (same path as Gemini).

3. generic_openai_provider.py — max_tokens raised 1024→4096 (classify) / 256→1024 (suggest)
   Qwen3 thinking models (like qwen/qwen3.5-9b) consume reasoning tokens from the
   same budget. With max_tokens=1024 the model exhausts the budget in the thinking
   trace, leaving content=''. 4096 gives room for both thinking and the JSON output.

Also updates lmstudio PROVIDER_DEFAULTS model to qwen/qwen3.5-9b (confirmed loaded
in LM Studio) and context_chars to 32_000 (Qwen3 context window).

Tested: classify(invoice_text, ['Finance','Legal','HR']) → topics=['Finance'] ✓
Backend tests: 376 passed ✓

Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>

2026-06-05 00:00:26 +02:00

curo1305

02bcbb9143

feat(07-02): singleton OpenAIProvider + GenericOpenAIProvider + MAX_AI_CHARS removal

- openai_provider.py: singleton self._client=AsyncOpenAI(...) in __init__ (D-07),
  _truncate() 60/40 smart truncation (D-13), removed MAX_AI_CHARS constant,
  changed __init__ signature to include context_chars, removed _client() method
- generic_openai_provider.py: new class, subclasses OpenAIProvider, conditional
  response_format={"type":"json_object"} on supports_json_mode flag (D-01/D-02),
  imports parse_classification + parse_suggestions from ai.utils (D-02 contract)
- ollama_provider.py: added context_chars kwarg with default 8000, passes through
- lmstudio_provider.py: added context_chars kwarg with default 8000, passes through
- classifier.py: removed MAX_AI_CHARS constant and text[:MAX_AI_CHARS] slices;
  truncation now handled inside each provider via _truncate()
- requirements.txt: bumped anthropic floor to >=0.95.0 (D-03 output_config support)

2026-06-04 18:58:19 +02:00