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

Add the SCL resample TODO

This commit is contained in:
Edresson 2021-11-23 13:37:14 -03:00 committed by Eren Gölge
parent 4a8c344fc8
commit be8f444636
1 changed files with 3 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
TTS/tts/models/vits.py
View File

@ -409,6 +409,7 @@ class Vits(BaseTTS):
hasattr(self.speaker_encoder, "audio_config") hasattr(self.speaker_encoder, "audio_config")
and self.config.audio["sample_rate"] != self.speaker_encoder.audio_config["sample_rate"] and self.config.audio["sample_rate"] != self.speaker_encoder.audio_config["sample_rate"]
): ):
# TODO: change this with torchaudio Resample
raise RuntimeError( raise RuntimeError(
' [!] To use the speaker consistency loss (SCL) you need to have matching sample rates between the TTS model ({}) and the speaker encoder ({})!' ' [!] To use the speaker consistency loss (SCL) you need to have matching sample rates between the TTS model ({}) and the speaker encoder ({})!'
.format(self.config.audio["sample_rate"], self.speaker_encoder.audio_config["sample_rate"]) .format(self.config.audio["sample_rate"], self.speaker_encoder.audio_config["sample_rate"])
@ -418,8 +419,8 @@ class Vits(BaseTTS):
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