forked from mrq/tortoise-tts
Fixed silly lack of EOF blank line, indentation
This commit is contained in:
parent
5a74461c1e
commit
0e08760896
|
@ -81,13 +81,13 @@ if __name__ == '__main__':
|
||||||
dbg_state = (seed, texts, voice_samples, conditioning_latents)
|
dbg_state = (seed, texts, voice_samples, conditioning_latents)
|
||||||
torch.save(dbg_state, f'debug_states/read_debug_{selected_voice}.pth')
|
torch.save(dbg_state, f'debug_states/read_debug_{selected_voice}.pth')
|
||||||
|
|
||||||
# Combine each candidate's audio clips.
|
# Combine each candidate's audio clips.
|
||||||
if args.candidates > 1:
|
if args.candidates > 1:
|
||||||
audio_clips = []
|
audio_clips = []
|
||||||
for candidate in range(args.candidates):
|
for candidate in range(args.candidates):
|
||||||
for line in range(len(texts)):
|
for line in range(len(texts)):
|
||||||
wav_file = os.path.join(voice_outpath, str(line), f"{candidate}.wav")
|
wav_file = os.path.join(voice_outpath, str(line), f"{candidate}.wav")
|
||||||
audio_clips.append(load_audio(wav_file, 24000))
|
audio_clips.append(load_audio(wav_file, 24000))
|
||||||
audio_clips = torch.cat(audio_clips, dim=-1)
|
audio_clips = torch.cat(audio_clips, dim=-1)
|
||||||
torchaudio.save(os.path.join(voice_outpath, f"combined_{candidate:02d}.wav"), audio_clips, 24000)
|
torchaudio.save(os.path.join(voice_outpath, f"combined_{candidate:02d}.wav"), audio_clips, 24000)
|
||||||
audio_clips = []
|
audio_clips = []
|
||||||
|
|
Loading…
Reference in New Issue
Block a user