This commit is contained in:
Yuta Hayashibe 2022-09-24 14:22:33 +09:00
parent 2a159f12d7
commit ca7e77f437

View file

@ -199,9 +199,9 @@ class WhisperStreamingTranscriber:
self.timestamp += last_timestamp_position0 * self.time_precision self.timestamp += last_timestamp_position0 * self.time_precision
yield last_timestamp_position0 yield last_timestamp_position0
else: else:
logger.debug("Length of consecutive: 0")
duration = segment_duration duration = segment_duration
timestamps = tokens[timestamp_tokens.nonzero().flatten()] timestamps = tokens[timestamp_tokens.nonzero().flatten()]
logger.debug(f"Length of consecutive: 0, timestamps: {timestamps}")
if len(timestamps) > 0: if len(timestamps) > 0:
# no consecutive timestamps but it has a timestamp; use the last one. # no consecutive timestamps but it has a timestamp; use the last one.
# single timestamp at the end means no speech after the last timestamp. # single timestamp at the end means no speech after the last timestamp.
@ -209,6 +209,7 @@ class WhisperStreamingTranscriber:
timestamps[-1].item() - self.tokenizer.timestamp_begin timestamps[-1].item() - self.tokenizer.timestamp_begin
) )
duration = last_timestamp_position * self.time_precision duration = last_timestamp_position * self.time_precision
logger.debug(f"segment_duration: {segment_duration}, Duration: {duration}")
chunk = self._get_chunk( chunk = self._get_chunk(
start=self.timestamp, start=self.timestamp,
end=self.timestamp + duration, end=self.timestamp + duration,
@ -281,7 +282,7 @@ class WhisperStreamingTranscriber:
seek += segment.shape[-1] seek += segment.shape[-1]
rest_start = None rest_start = None
else: else:
seek += last_timestamp_position seek += last_timestamp_position * self.input_stride
rest_start = seek rest_start = seek
logger.debug(f"Last rest_start={rest_start}") logger.debug(f"Last rest_start={rest_start}")