PyTorch_Speaker_Verification icon indicating copy to clipboard operation
PyTorch_Speaker_Verification copied to clipboard

Published 20 hours ago • verified publisher icon HarryVolek

Very slow training on CPU

Open chrisspen opened this issue 5 years ago • 2 comments

I'm training with the CPU backend for Torch with Python3.7 on anaconda with a quad-core CPU, and it seems very slow. After running for 12 hours, it had only completely 30 epoches out of the max of 950, and my loss had only gotten down to about 12. I estimate a full training would likely take an entire week. Is this normal without cuda? Is there anything I can do to speed up performance without getting a GPU?

chrisspen avatar Oct 14 '19 19:10 chrisspen

Hi @chrisspen , I had the same issue here, did you found a solution ? Could you please tell me how did you solve it?

MahmoudAliEng avatar May 08 '20 19:05 MahmoudAliEng

Yes, that is normal with big models. You need a GPU.

fmaker avatar May 31 '20 23:05 fmaker