auto-save 2026-05-17 13:23 (~2)
This commit is contained in:
@@ -43,6 +43,7 @@ LLM_BASE_URL = os.getenv("LLM_BASE_URL", "").strip()
|
||||
LLM_API_KEY = os.getenv("LLM_API_KEY", "").strip()
|
||||
ASR_MODEL = os.getenv("ASR_MODEL", "whisper-1")
|
||||
ASR_FALLBACK_MODEL = os.getenv("ASR_FALLBACK_MODEL", "gemini-2.5-flash").strip() or "gemini-2.5-flash"
|
||||
ASR_TIMEOUT_SECONDS = max(15, int(os.getenv("ASR_TIMEOUT_SECONDS", "45")))
|
||||
TRANSLATE_MODEL = os.getenv("TRANSLATE_MODEL", "gemini-2.5-flash")
|
||||
REWRITE_MODEL = os.getenv("REWRITE_MODEL", "gemini-2.5-pro")
|
||||
VISION_MODEL = os.getenv("VISION_MODEL", "gemini-2.5-flash")
|
||||
@@ -1752,6 +1753,7 @@ def _transcribe_gemini_sync(wav: Path) -> list[dict]:
|
||||
{"type": "input_audio", "input_audio": {"data": audio_b64, "format": "wav"}},
|
||||
]}],
|
||||
temperature=0,
|
||||
timeout=ASR_TIMEOUT_SECONDS,
|
||||
)
|
||||
content = (resp.choices[0].message.content or "").strip()
|
||||
return _parse_asr_segments(content, duration)
|
||||
@@ -1771,6 +1773,7 @@ def _transcribe_sync(wav: Path) -> list[dict]:
|
||||
model=ASR_MODEL,
|
||||
response_format="verbose_json",
|
||||
timestamp_granularities=["segment"],
|
||||
timeout=ASR_TIMEOUT_SECONDS,
|
||||
)
|
||||
raw = resp.model_dump() if hasattr(resp, "model_dump") else resp
|
||||
segments = raw.get("segments") or []
|
||||
|
||||
Reference in New Issue
Block a user