Generated audio has no voice

CSTR-Edinburgh / merlin

This is now the official location of the Merlin project.

http://www.cstr.ed.ac.uk/projects/merlin/

Apache License 2.0

1.31k stars 440 forks source link

Generated audio has no voice #457

Open mukundroy opened 5 years ago

mukundroy commented 5 years ago

Any idea why the generated audio has no voice. I trained the model with 10K utterances and training completed successfully also. I am attaching sample generated audio. Kindly give insight what wrong could have gone?

audio.zip

mirfan899 commented 5 years ago

Same issue here. I trained the model on GPU and synthesized files but these files have no voice.

RasmusD commented 5 years ago

It would be much easier if you provided the log from training/synthesis to give any opinion on what may be wrong.

priyanthini commented 4 years ago

Got the same problem. Did anyone find a solution for it?

RasmusD commented 4 years ago

Please attach your training and/or synthesis log to allow for anybody to help

priyanthini commented 4 years ago

feed_forward_6_tanh_12_52PM_February_08_2020.log feed_forward_6_tanh_11_30AM_February_09_2020.log

@RasmusD herewith I have attached acoustic model log