Mixed precision training

TRI-ML / packnet-sfm

TRI-ML Monocular Depth Estimation Repository

https://tri-ml.github.io/packnet-sfm/

MIT License

1.21k stars 241 forks source link

Mixed precision training #231

Open hemangchawla opened 2 years ago

hemangchawla commented 2 years ago

Hello,

Do you have an example on how to use mixed precision training for smaller GPU memory? I see that apex is part of the Dockerfile. However, I am not sure where the amp is used.

Could you please assist?

Kind regards