Medix-AI/coqui-tts
Medix-AI
/
coqui-tts
mirror of https://github.com/coqui-ai/TTS.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

Fix pylint checks

This commit is contained in:
Edresson 2021-11-23 11:43:08 -03:00 committed by Eren Gölge
parent 4cd0e4eb0d
commit 78a23e19df
1 changed files with 11 additions and 8 deletions
TTS/tts/models
vits.py

11
TTS/tts/models/vits.py
View File

@ -412,12 +412,14 @@ class Vits(BaseTTS):
raise RuntimeError( raise RuntimeError(
" [!] To use the speaker consistency loss (SCL) you need to have the TTS model sampling rate ({}) equal to the speaker encoder sampling rate ({}) !".format(self.audio_config["sample_rate"], self.speaker_encoder.audio_config["sample_rate"]) " [!] To use the speaker consistency loss (SCL) you need to have the TTS model sampling rate ({}) equal to the speaker encoder sampling rate ({}) !".format(self.audio_config["sample_rate"], self.speaker_encoder.audio_config["sample_rate"])
) )
'''self.audio_transform = torchaudio.transforms.Resample( # pylint: disable=W0101,W0105
""" self.audio_transform = torchaudio.transforms.Resample(
orig_freq=self.audio_config["sample_rate"], orig_freq=self.audio_config["sample_rate"],
new_freq=self.speaker_encoder.audio_config["sample_rate"], new_freq=self.speaker_encoder.audio_config["sample_rate"],
) )
else: else:
self.audio_transform = None''' self.audio_transform = None
"""
else: else:
# self.audio_transform = None # self.audio_transform = None
self.speaker_encoder = None self.speaker_encoder = None
@ -657,8 +659,9 @@ class Vits(BaseTTS):
wavs_batch = torch.cat((wav_seg, o), dim=0).squeeze(1) wavs_batch = torch.cat((wav_seg, o), dim=0).squeeze(1)
# resample audio to speaker encoder sample_rate # resample audio to speaker encoder sample_rate
'''if self.audio_transform is not None: # pylint: disable=W0105
wavs_batch = self.audio_transform(wavs_batch)''' """if self.audio_transform is not None:
wavs_batch = self.audio_transform(wavs_batch)"""
pred_embs = self.speaker_encoder.forward(wavs_batch, l2_norm=True) pred_embs = self.speaker_encoder.forward(wavs_batch, l2_norm=True)