olaf2k / neural_tts Goto Github PK

i am not entirely sure if i understood your description right. the script expects many files in input folder, witch each file containing one sentence without newlines?

split3.txt

for example if i put both sentences in separate files in input...i get in output two folders with each containing one wav. the wavs are correct.

if i put both sentences into one file, with each in a newline...only one of them gets processed

if i do like in the example above, two sentences without newline, i get one folder in output, cotnaining 2 wavs. the first is correct, but the last sentences always the last word gets repeated multiple times until timeout.

can you please clarify the expected input?

i would like to change tacotron to speedy-speech-wn tts model in the script. any chance you can point me on where i need to change that ? ( i did not find the string directly in the script)

the model is listed when i type tts --list_models and i find it better sounding than tacotron2