Mirroristas/whisper
2
0
Fork 0
mirror of https://github.com/openai/whisper.git synced 2025-11-24 14:35:57 +00:00
Code Issues Packages Projects Releases Wiki Activity

detect language based on available frames not seek

Browse Source
This commit is contained in:
Kent Slaney 2024-07-22 13:40:12 -07:00
parent 247391a2af
commit 092cb3409e
1 changed files with 6 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
whisper/transcribe.py
View File

@ -122,6 +122,7 @@ class Transcriber(metaclass=PassthroughPropertyDefaults):
_hypothesis: LanguageHypothesis = LanguageHypothesis() _hypothesis: LanguageHypothesis = LanguageHypothesis()
_language: Optional[str] _language: Optional[str]
_language_detection_warned: bool = False
@PassthroughProperty[Optional[str]](None).property @PassthroughProperty[Optional[str]](None).property
def language(self) -> Optional[str]: def language(self) -> Optional[str]:
@ -129,15 +130,18 @@ class Transcriber(metaclass=PassthroughPropertyDefaults):
return self._language return self._language
if not self.model.is_multilingual: if not self.model.is_multilingual:
return "en" return "en"
if self.verbose: if self.verbose and not self._language_detection_warned:
print( print(
"Detecting language using up to the first 30 seconds." "Detecting language using up to the first 30 seconds."
"Use `--language` to specify the language" "Use `--language` to specify the language"
) )
self._language_detection_warned = True
if self.latest is None: if self.latest is None:
return None return None
if self._seek == self._hypothesis.last: available = self.frame_offset + self.latest.shape[-1]
if available == self._hypothesis.last:
return self._hypothesis.language return self._hypothesis.language
self._hypothesis.last = available
if self.frame_offset > 0 or self.latest.shape[-1] == N_FRAMES * 2: if self.frame_offset > 0 or self.latest.shape[-1] == N_FRAMES * 2:
mel = ( mel = (
self.latest self.latest
@ -146,7 +150,6 @@ class Transcriber(metaclass=PassthroughPropertyDefaults):
) )
self._language = self.detect_language(mel) self._language = self.detect_language(mel)
return self._language return self._language
self._hypothesis.last = self._seek or 0
self._hypothesis.since += 1 self._hypothesis.since += 1
if 2**self._hypothesis.evidence > self._hypothesis.since: if 2**self._hypothesis.evidence > self._hypothesis.since:
return self._hypothesis.language return self._hypothesis.language