How many GPUs or Cards are used when training?

shariqfarooq123 / AdaBins

Official implementation of Adabins: Depth Estimation using adaptive bins

GNU General Public License v3.0

727 stars 156 forks source link

Closed chenerg closed 3 years ago

chenerg commented 3 years ago

I have noticed that you use syncbn in your code.

shariqfarooq123 commented 3 years ago

Final models were trained on a single node with 4 V100 GPUs.