Tacotron2: Please explain the Setup again elaborately

[deadskull7]

This is to the contributors of this repository and any other 3rd person who knows how to setup the repo. in one's machine till fully working and result stage. I am having some problem in 4th part of setup. Please help me out. Thanks in advance.

Update .wav paths: sed -i -- 's,DUMMY,ljs_dataset_folder/wavs,g' filelists/*.txt

    Alternatively, set load_mel_from_disk=True in hparams.py and update mel-spectrogram paths

Its coming like 'sed' is not recognized as an internal or external command, operable program or batch file.

[deadskull7]

My Bad .... Its done now. Used ubuntu .

[EuphoriaCelestial]

please explain it again, I dont know what need to be changed accordingly

[CookiePPP]

@EuphoriaCelestial
the command will replace all DUMMY lines inside the filelists folder with ljs_dataset_folder/wavs.

Replace ljs_dataset_folder inside

sed -i -- 's,DUMMY,ljs_dataset_folder/wavs,g' filelists/*.txt

with the path to your LJSpeech-1.1 directory.

EDIT: This assumes your using Linux. Reply if you need Windows instructions. Windows is more complicated.

[EuphoriaCelestial]

@CookiePPP thank you! since the command only correct the file path in the filelists folder, I have done it using Text editor on Ubuntu. It will work right? No issue later on?

[CookiePPP]

@EuphoriaCelestial
That should work yes.