search

mkotha / WaveRNN

A WaveRNN implementation
MIT License
198 stars 48 forks source link

How to lower memory usage of Cuda? #14

Open raanubis opened 4 years ago

raanubis commented 4 years ago

Hi,

thanks for the nice work.

Preprocess is already finished fo vqvae model (multispeaker) I have a dataset size with 498005 files. If I try to use "wavernn.py -m vqvae" I get the error message that my Cuda memory run out. RuntimeError: CUDA out of memory. Tried to allocate 1.70 GiB (GPU 0; 8.00 GiB total capacity; 5.08 GiB already allocated; 929.97 MiB free; 27.78 MiB cached)

How it is possible (which option I need to set) that it's run with my card?

I own only a graphic card with 8 GB

thanks