search

WongKinYiu / yolor

implementation of paper - You Only Learn One Representation: Unified Network for Multiple Tasks (https://arxiv.org/abs/2105.04206)
GNU General Public License v3.0
1.99k stars 521 forks source link

minimum GPU VRAM to use YOLOR #177

Closed rodthebuilder1 closed 2 years ago

rodthebuilder1 commented 2 years ago

Hi,

I am limited to a rtx 2070 super with 8GB of VRAM, all on windows 10

I was wondering if it would be possible to run YOLOR on it ? OR would I need more Vram? Is it possible to use cpu RAM to complement VRAM ?

Thanks

susanin1970 commented 2 years ago

I tested the YOLOR-CSP-X inference on a single frame on a GTX 1080 Ti with 11Gb of VRAM Inferaece was successful

atamazian commented 2 years ago

Your question is quite vague. First of all, do you want to train models, or you just want to do inference? Inference requires less GPU VRAM.