Closed marshonhuckleberry closed 4 years ago
i was confused
inference.py -p [checkpoint path] -i [input mel path]
is actually
inference.py -p [checkpoint path] -i [input mel dir path]
nvidia_tacotron2_LJ11_epoch6400.pt is the checkpoint
each mel file has ".mel" extension not ".pt" as i thought
don't know yet to output mel from tacotron2
but i found some mel ".pt" files in nvidia waveglow repo
renamed them to ".mel" and it worked
inference.py -p [checkpoint path] -i [input mel path]
wich files i use to generate speech? found this file "nvidia_tacotron2_LJ11_epoch6400.pt", what is it for?