admin-valentin/WhisperLiveKit
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

chunk at sentence takes now an argument =self.comited

Browse source
This commit is contained in:
Silas Kieser 2025-01-27 16:29:06 +01:00
parent 04170153e0
commit 77d43885a3
1 changed files with 10 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

17
src/whisper_streaming/online_asr.py
View file

@ -173,9 +173,11 @@ class OnlineASRProcessor:
self.transcript_buffer.insert(tsw, self.buffer_time_offset) self.transcript_buffer.insert(tsw, self.buffer_time_offset)
o = self.transcript_buffer.flush() o = self.transcript_buffer.flush()
# Completed words
self.commited.extend(o) self.commited.extend(o)
completed = self.to_flush(o) completed = self.to_flush(o)
logger.debug(f">>>>COMPLETE NOW: {completed[2]}") logger.debug(f">>>>COMPLETE NOW: {completed[2]}")
## The rest is incomplete
the_rest = self.to_flush(self.transcript_buffer.complete()) the_rest = self.to_flush(self.transcript_buffer.complete())
logger.debug(f"INCOMPLETE: {the_rest[2]}") logger.debug(f"INCOMPLETE: {the_rest[2]}")
@ -183,11 +185,12 @@ class OnlineASRProcessor:
if self.buffer_trimming_way == "sentence": if self.buffer_trimming_way == "sentence":
self.chunk_completed_sentence() self.chunk_completed_sentence(self.commited)
# TODO: new words in `completed` should not be reterned unless they form a sentence
# TODO: only complete sentences should go to completed
if len(self.audio_buffer) / self.SAMPLING_RATE > self.buffer_trimming_sec : if len(self.audio_buffer) / self.SAMPLING_RATE > self.buffer_trimming_sec :
@ -219,13 +222,13 @@ class OnlineASRProcessor:
return self.to_flush(o) return completed
def chunk_completed_sentence(self): def chunk_completed_sentence(self, commited_text):
if self.commited == []: if commited_text == []:
return return
sents = self.words_to_sentences(self.commited) sents = self.words_to_sentences(commited_text)
@ -436,7 +439,7 @@ class VACOnlineASRProcessor(OnlineASRProcessor):
ret = self.online.process_iter() ret = self.online.process_iter()
return ret return ret
else: else:
print("no online update, only VAD", self.status, file=self.logfile) logger.debug("no online update, only VAD")
return (None, None, "") return (None, None, "")
def finish(self): def finish(self):