voice conversion error ['NoneType' object has no attribute 'float'] on colab

[bnsantoso]

failed to generate audio when using voice conversion using subtoaudio coqui wraper.

logs :

[/usr/local/lib/python3.10/dist-packages/subtoaudio/subtoaudio.py](https://localhost:8080/#) in convert_to_audio(self, sub_data, speaker, language, voice_conversion, speaker_wav, voice_dir, output_path, tempo_mode, tempo_speed, tempo_limit, shift_mode, shift_limit, save_temp, speed, emotion, **kwargs)
    120       for entry_data in data:
    121         audio_path = f"{temp_folder}/{entry_data['audio_name']}"
--> 122         self.apitts.tts_with_vc_to_file(f"{entry_data['text']}",file_path=audio_path,**convert_param)
    123 
    124 

[/usr/local/lib/python3.10/dist-packages/TTS/api.py](https://localhost:8080/#) in tts_with_vc_to_file(self, text, language, speaker_wav, file_path)
    473                 Output file path. Defaults to "output.wav".
    474         """
--> 475         wav = self.tts_with_vc(text=text, language=language, speaker_wav=speaker_wav)
    476         save_wav(wav=wav, path=file_path, sample_rate=self.voice_converter.vc_config.audio.output_sample_rate)

[/usr/local/lib/python3.10/dist-packages/TTS/api.py](https://localhost:8080/#) in tts_with_vc(self, text, language, speaker_wav)
    451         if self.voice_converter is None:
    452             self.load_vc_model_by_name("voice_conversion_models/multilingual/vctk/freevc24")
--> 453         wav = self.voice_converter.voice_conversion(source_wav=fp.name, target_wav=speaker_wav)
    454         return wav
    455 

[/usr/local/lib/python3.10/dist-packages/TTS/utils/synthesizer.py](https://localhost:8080/#) in voice_conversion(self, source_wav, target_wav)
    251 
    252     def voice_conversion(self, source_wav: str, target_wav: str) -> List[int]:
--> 253         output_wav = self.vc_model.voice_conversion(source_wav, target_wav)
    254         return output_wav
    255 

[/usr/local/lib/python3.10/dist-packages/torch/utils/_contextlib.py](https://localhost:8080/#) in decorate_context(*args, **kwargs)
    113     def decorate_context(*args, **kwargs):
    114         with ctx_factory():
--> 115             return func(*args, **kwargs)
    116 
    117     return decorate_context

[/usr/local/lib/python3.10/dist-packages/TTS/vc/models/freevc.py](https://localhost:8080/#) in voice_conversion(self, src, tgt)
    645         """
    646 
--> 647         wav_tgt = self.load_audio(tgt).cpu().numpy()
    648         wav_tgt, _ = librosa.effects.trim(wav_tgt, top_db=20)
    649 

[/usr/local/lib/python3.10/dist-packages/TTS/vc/models/freevc.py](https://localhost:8080/#) in load_audio(self, wav)
    630         if isinstance(wav, list):
    631             wav = torch.from_numpy(np.array(wav)).to(self.device)
--> 632         return wav.float()
    633 
    634     @torch.inference_mode()

AttributeError: 'NoneType' object has no attribute 'float'

[bnsantoso]

use this https://github.com/bnsantoso/sub-to-audio#voice-conversion instead using colab form.