|
@@ -95,7 +95,7 @@ class PretsselVocoderAgent(NoUpdateTargetMixin, TextToSpeechAgent):
|
|
|
|
|
|
return WriteAction(
|
|
|
SpeechSegment(
|
|
|
- content=wav[0][0].tolist(),
|
|
|
+ content=wav[0][0][0].tolist(),
|
|
|
finished=states.source_finished,
|
|
|
sample_rate=self.vocoder_sample_rate,
|
|
|
tgt_lang=tgt_lang,
|