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

Prevent resampling

This commit is contained in:
Eren G??lge 2023-11-05 00:43:28 +01:00
parent 2d65f005a2
commit 9b5c295fcf
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/xtts.py
View File

@ -377,8 +377,8 @@ class Xtts(BaseTTS):
sr (int): Sample rate of the audio. sr (int): Sample rate of the audio.
length (int): Length of the audio in seconds. Defaults to 3. length (int): Length of the audio in seconds. Defaults to 3.
""" """
if sr != 22050:
audio_22k = torchaudio.functional.resample(audio, sr, 22050) audio_22k = torchaudio.functional.resample(audio, sr, 22050)
audio_22k = audio_22k[:, : 22050 * length] audio_22k = audio_22k[:, : 22050 * length]
if self.args.gpt_use_perceiver_resampler: if self.args.gpt_use_perceiver_resampler:
mel = wav_to_mel_cloning(audio_22k, mel = wav_to_mel_cloning(audio_22k,
@ -600,6 +600,7 @@ class Xtts(BaseTTS):
(gpt_cond_latent, diffusion_conditioning, speaker_embedding) = self.get_conditioning_latents( (gpt_cond_latent, diffusion_conditioning, speaker_embedding) = self.get_conditioning_latents(
audio_path=ref_audio_path, gpt_cond_len=gpt_cond_len, max_ref_length=max_ref_len, sound_norm_refs=sound_norm_refs audio_path=ref_audio_path, gpt_cond_len=gpt_cond_len, max_ref_length=max_ref_len, sound_norm_refs=sound_norm_refs
) )
return self.inference( return self.inference(
text, text,
language, language,