Closed AlucardNosferatu closed 4 years ago
Hi, I have a question about requirements of this project, I've noticed that fast inference wavenet kernels are used in this project, do we need a GPU to run this program? Is it possible to output generated result with only CPU? Thank you.
You can use sample_py.sh to run pure-python sampling. Be prepared for slow inference times (minutes per second of audio generated)
sample_py.sh
Hi, I have a question about requirements of this project, I've noticed that fast inference wavenet kernels are used in this project, do we need a GPU to run this program? Is it possible to output generated result with only CPU? Thank you.