heimoshuiyu/faster-whisper
1
0
Fork 0
Code Issues Pull Requests 1 Packages Projects Releases Wiki Activity

Added retrieval mechanism (avg_log_prob/no_speech_prob) (#103)

Browse Source 
* Added retrieval mechanism 

Added retrieval mechanism to retrieve avg_log_prob and no_speech_prob from the Transcribe method.

* Update transcribe.py

* Update transcribe.py

* Initial commit
This commit is contained in:
palladium123
2023-04-03 22:56:35 +08:00
committed by GitHub
parent f20bb258de
commit b4c1c57781
1 changed files with 4 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
faster_whisper/transcribe.py
View File

@@ -26,6 +26,8 @@ class Segment(NamedTuple):
end: float end: float
text: str text: str
words: Optional[List[Word]] words: Optional[List[Word]]
avg_log_prob: float
no_speech_prob: float
class AudioInfo(NamedTuple): class AudioInfo(NamedTuple):
@@ -441,6 +443,8 @@ class WhisperModel:
if options.word_timestamps if options.word_timestamps
else None else None
), ),
avg_log_prob=avg_log_prob,
no_speech_prob=result.no_speech_prob,
) )
def encode(self, features: np.ndarray) -> ctranslate2.StorageView: def encode(self, features: np.ndarray) -> ctranslate2.StorageView: