michellelychan commited on
Commit
467f7e2
1 Parent(s): f5c319c

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +15 -0
app.py CHANGED
@@ -19,9 +19,24 @@ processor = AutoProcessor.from_pretrained("suno/bark")
19
  voice_preset = "v2/en_speaker_3"
20
 
21
  # generate audio
 
 
 
 
 
 
 
 
 
22
  def generate_audio(text, preset, output_file_name="bark_generation"):
23
  file_name = output_file_name + ".wav"
24
  inputs = processor(text, voice_preset=preset)
 
 
 
 
 
 
25
  audio_array = model.generate(**inputs)
26
  audio_array = audio_array.cpu().numpy().squeeze()
27
  sample_rate = model.generation_config.sample_rate
 
19
  voice_preset = "v2/en_speaker_3"
20
 
21
  # generate audio
22
+ # def generate_audio(text, preset, output_file_name="bark_generation"):
23
+ # file_name = output_file_name + ".wav"
24
+ # inputs = processor(text, voice_preset=preset)
25
+ # audio_array = model.generate(**inputs)
26
+ # audio_array = audio_array.cpu().numpy().squeeze()
27
+ # sample_rate = model.generation_config.sample_rate
28
+ # write_wav(file_name, sample_rate, audio_array)
29
+ # return file_name
30
+
31
  def generate_audio(text, preset, output_file_name="bark_generation"):
32
  file_name = output_file_name + ".wav"
33
  inputs = processor(text, voice_preset=preset)
34
+
35
+ # Ensure the inputs are on the right device
36
+ for k, v in inputs.items():
37
+ if isinstance(v, torch.Tensor):
38
+ inputs[k] = v.to(device)
39
+
40
  audio_array = model.generate(**inputs)
41
  audio_array = audio_array.cpu().numpy().squeeze()
42
  sample_rate = model.generation_config.sample_rate