Init speaker manager for speaker encoder

This commit is contained in:
Eren Gölge 2021-12-22 15:51:53 +00:00
parent 633dcc9c56
commit 5c5ddd2ba7
2 changed files with 10 additions and 3 deletions

View File

@ -256,11 +256,11 @@ class SpeakerManager:
self.speaker_encoder.load_checkpoint(config_path, model_path, eval=True, use_cuda=self.use_cuda)
self.speaker_encoder_ap = AudioProcessor(**self.speaker_encoder_config.audio)
def compute_d_vector_from_clip(self, wav_file: Union[str, list]) -> list:
def compute_d_vector_from_clip(self, wav_file: Union[str, List[str]]) -> list:
"""Compute a d_vector from a given audio file.
Args:
wav_file (Union[str, list]): Target file path.
wav_file (Union[str, List[str]]): Target file path.
Returns:
list: Computed d_vector.

View File

@ -171,7 +171,12 @@ class Synthesizer(object):
def _init_speaker_encoder(self, speaker_manager):
"""Initialize the SpeakerEncoder"""
if self.encoder_checkpoint:
speaker_manager.init_speaker_encoder(self.encoder_checkpoint, self.encoder_config)
if speaker_manager is None:
speaker_manager = SpeakerManager(
encoder_model_path=self.encoder_checkpoint, encoder_config_path=self.encoder_config
)
else:
speaker_manager.init_speaker_encoder(self.encoder_checkpoint, self.encoder_config)
return speaker_manager
def _init_language_manager(self):
@ -183,6 +188,8 @@ class Synthesizer(object):
language_manager = LanguageManager(language_ids_file_path=self.tts_languages_file)
elif self.tts_config.get("language_ids_file", None):
language_manager = LanguageManager(language_ids_file_path=self.tts_config.language_ids_file)
else:
language_manager = LanguageManager(config=self.tts_config)
return language_manager
def _load_vocoder(self, model_file: str, model_config: str, use_cuda: bool) -> None: