Traceback (most recent call last):
  File "pl_one_lstm_lps_separate.py", line 126, in <module>
    separate(args)
  File "pl_one_lstm_lps_separate.py", line 115, in separate
    write(lstm_est_speech.cuda().data.cpu().detach().numpy(), filename + '_en.wav')
  File "pl_one_lstm_lps_separate.py", line 72, in write
    librosa.output.write_wav(filename, inputs, sr)# norm=True)
  File "<decorator-gen-10>", line 2, in write_wav
  File "/home/limiao/.conda/envs/Pytorch/lib/python3.7/site-packages/librosa/util/decorators.py", line 58, in __wrapper
    return func(*args, **kwargs)
  File "/home/limiao/.conda/envs/Pytorch/lib/python3.7/site-packages/librosa/output.py", line 239, in write_wav
    util.valid_audio(y, mono=False)
  File "/home/limiao/.conda/envs/Pytorch/lib/python3.7/site-packages/librosa/util/utils.py", line 275, in valid_audio
    raise ParameterError('Audio buffer is not finite everywhere')
librosa.util.exceptions.ParameterError: Audio buffer is not finite everywhere

问题出现是因为est_amplitude = torch.sqrt(torch.exp(lstm_out1) - 1.0)这行代码中torch.exp(lstm_out1) - 1.0会出现负数的情况，然后我又需要将est_amplitude结果写入音频，才导致出现这个Bug，所以微调因子还需要再调整！