search

r9y9 / nnmnkwii_gallery

A collection of examples demonstrating how we can build speech synthesis systems using nnmnkwii.
https://github.com/r9y9/nnmnkwii
Other
71 stars 21 forks source link

No speech was generated #3

Closed korguchi closed 6 years ago

korguchi commented 6 years ago

Thank you for your many responses. I have found another issue and report it. The process progressed smoothly, and it finally entered the stage of outputting sound. However, no speech was output in [29].

https://gist.github.com/KORGUCHI/a82927ceb5dbdfa5e7b7aa01701986a7

r9y9 commented 6 years ago

Seems there's no error. Have you checked if samples exist in generated directory?

r9y9 commented 6 years ago

I'm wondering IPython.display.display is not working correctly.

r9y9 commented 6 years ago

Umm, I missed your comment at https://github.com/r9y9/nnmnkwii_gallery/issues/2#issuecomment-399639418. That's weird. 

test_label_paths = sorted(glob(join(DATA_ROOT, "test_label_phone_align", "*.lab")))
hts_generated_wav_files = sorted(glob(join("./generated/hts_nit_atr503_2mix/*.wav")))

Please inspect the above variables at the last cell in the notebook to check if we have test inputs.

korguchi commented 6 years ago

I found that there were no wav files in ./generated/hts_nit_atr503_2mix, so after giving the filenames into hts_generated_wav_files, it outputted the speeches. I'm sorry for my trivial mistake and appreciate your help.

r9y9 commented 6 years ago

Ah sorry. That's my fault. I haven't pushed the directory to the repo. I will do that immediately.

r9y9 commented 6 years ago

Fixed.

korguchi commented 6 years ago

I can confirm it has been fixed and get the same result of the default. Thanks a lot.